IT Svit

75% AWS cost and resource optimization in just one week for Shoebacca

Learn how IT Svit helped Shoebacca’s system survive peak loads during the holidays and remain stable throughout, as well as reduce AWS costs by 75%.

 Download PDF

Christmas and New Year are some of the most heartwarming holidays that bring a ton of joy and happiness, everyone looks forward to. Or so we believe. While this is definitely true for most of us, holiday seasons can also put a lot of pressure on businesses worldwide. However, Christmas is all about miracles, right? Well, let us tell you how IT Svit made a marvel for one amazing online retailer – Shoebacca.

Shoebacca: What do they do?

Shoebacca is a Texas-based private company that offers footwear and apparel of all different sizes and styles. On the website, you can shop for brands such as Puma, Keen, Saucony, Reebok, and many others. With Shoebacca, you’ll find a pair to your liking. The company aims to provide its clients with hassle-free shopping at a reasonable price. 

Challenge details

As our client approached the season of Christmas and New Year, they experienced increased concern regarding the ability of infrastructure to handle an upcoming surge in traffic. These concerns emerged due to the issue occurring regularly – the client’s existing setup would collapse without apparent reason, and the client anticipated that the same thing would happen again during the holiday season. 

A robust and reliable infrastructure is especially critical because of a growing demand for online shopping and other digital services. Many websites crash when experiencing peak traffic if they are not prepared for that sort of system overload, so it’s crucial to ensure that the entire infrastructure can withstand the surge in traffic during those peak times. 

Technical roadmap

1. Reconfigure volume snapshot lifecycle policy, as there is currently no proper rotation, with 922 snapshots in total. Moreover, snapshots are done every 2–4 hours, which is excessive and not needed. Lifecycle policy also requires some touch–ups, so we plan to reconfigure it as well. To do that, we will change the daily and weekly snapshots schedule.
2. Set up CloudWatch agents. Currently, CPU/Memory metrics from Windows hosts are not collected, which prevents us from analyzing data in the long term – it is impossible to get an analysis over at least one month.
3. Remove redundant volumes. Eight volumes are not in use – we will perform a backup and remove those volumes for now. 
4. We have provisioned a significant amount of IOPS for some volumes that are not currently being used. Our task is to decrease the provisioned IOPS while ensuring we avoid causing downtime during this process.

Our solutions

1. Identify the underlying causes of the platform’s vulnerability and come up with the most effective solutions within the limited timeframe, which is up to five working days. 
2. Optimize the cost of the current setup to save the budget but still have an opportunity to upgrade the whole project.
3. Take an active part in the project upgrading process and prepare cloud architecture design to make the updated project robust, cost-effective, easily scaled, and secure. 
4. Help design and build a high-volume event-driven messaging integration infrastructure for distributed systems.

Our outcomes

In just one week, the IT Svit team was able to achieve the following results: 

• The system successfully went through peak load during the holidays and remained stable throughout the period. 
• All resources were optimized.
• The infrastructure was designed cost-effectively and securely. 
• Snapshot policies were completely overhauled, and IOPSes decreased, reducing 75% of AWS cost (from $26000/month to $4000/month).

Alongside completing all objectives, we’ve partnered with Shoebacca to help them build a high-volume event-driven messaging integration infrastructure for their distributed systems.

Currently, we are actively participating in the early stages of the development process – we have already assisted the company’s in-house team with the containerization of the project and designed infrastructure cost-effectively and securely that provides the system with scalability and stability 24/7. More is to come – further infrastructure implementation is already underway!

A few last words: There is nothing impossible when you team up with professionals!

AI Text Recognition Solution for European HR Startup

Executive summary

Client: HURMA Systems, a SaaS company specializing in HR and recruitment automation.

Challenge: The primary challenge faced by HURMA was the manual data processing and transfer of thousands of resumes, complicated by the lack of standardized formats, arbitrary text structures, and multiple languages.

Solution: IT Svit implemented an AI-powered text recognition algorithm, revolutionizing the resume processing experience. This innovation enabled users to send resumes in various formats and languages effortlessly, saving time and optimizing their workflow.

HURMA: What Do They Do?

HURMA Systems is a SaaS company for automating HR and recruitment processes. Initially, we created the system for the company’s internal needs, so we used our extensive experience and expertise in DevOps, Big Data, and data security. We put our hearts and souls into creating HURMA Systems, and in 2018, it was launched to the public. Since then, it’s been transforming the approach to HR and recruiting in the European market.

Challenge details

One of the biggest challenges for SaaS companies is constantly optimizing their functionality to retain current customers and attract new ones. 

In the case of HURMA, the main audience for the product is recruiters. One of the frequent problems they face is the need to process and transfer data from hundreds of different resumes manually. 

It is also worth highlighting other issues that created extra obstacles for recruiters: 

1. There is no unified document format for a CV, so recruiters can receive PDF, DOCS, or even an IMG file. 
2. The text formatting is also arbitrary. For example, even in the name section, it can be a different order of name and surname. 
3. Users from multilingual countries also face the problem of receiving resumes in different languages.

Therefore, the HURMA team sought to provide users with a convenient solution to this problem.

Solution Implementation

IT Svit has extensive experience in implementing AI for business, so in this case, our team used this technology as the basis of the solution. The ultimate goal was to create a service for recognizing resumes’ text and transferring data to the internal functionality of HURMA.

To implement this idea, our engineers tested several text recognition algorithms. We tested the NER solution from NLTK, CoreNLP, and different off-the-shelf solutions, but such algorithms were unsuitable for searching for Slavic names. For our final product, we used a solution for Google library using the Celery library to improve stability under high loads.

As we mentioned earlier, all resumes have different formats, text patterns, and even languages, so the model had to be trained to work effectively in such conditions. A series of tests and improvements took 2 months to complete.

Now, users can send one or more resumes in different formats: PDF, TXT, DOCS, IMG, etc., and receive standardized data in HURMA. Also, if a person responds to a vacancy on a job portal and sends a resume, the service processes it and transfers the data to the system.

Such a solution allows users of HURMA to save hours of their time and optimize their workflow. 

Conclusion 

IT Svit has developed an AI solution for a leading European HR Tech startup. Our engineers trained the algorithm and implemented it in the HURMA environment so that users without technical skills could easily use it.

The solution has brought many benefits to our client, including increased product satisfaction among current customers and increased sales to new users.

Would you also like to boost your SaaS or implement an AI solution for business? IT Svit can become a reliable partner for your technology solutions.

Our client, the medical services application owner, turned to the IT Svit team with the request to transfer their medical app to the cloud and automate & simplify the file conversion for the further data analysis process. 

Case info

Location: France

Partnership period: 2018-ongoing

Team size: 2 – 4 people

Team location: Kharkiv, Ukraine

Services: cloud architecture, cloud infrastructure management, CI/CD

Expertise delivered: GCP cloud administration, DevOps services, cloud infrastructure management

Technologies: GCP, Kubernetes, Nginx, Jenkins, Ansible, Terraform.

Project requirements

The IT Svit team distinguished the following tasks:

• to design the new infrastructure on AWS or other providers;

• to ensure the system operational stability and resilience;

• to enable CI/CD pipelines for the file conversion processes;

• to enable simple system configuration and monitoring.

Challenges and Our solutions 

Challenges: 

• app deployment must be quicker and error-proof through CI/CD;

• app must be able to receive and store .raw files from external sources;

• .raw files must be converted into .zip files through a complex process;

• storage must be with folders for the ready files.

Our solutions: 

• to use a Google Cloud provider;

• to use Jenkins pipelines for the conversion of the files;

• to apply Ansible to install Jenkins, Docker, Java, Nginx, pip;

• to apply Terraform manifests to manage the future infrastructure in the cloud and to create the bucket in Google Cloud Storage for files;

• to use Nginx as a proxy server.

Results

Our work resulted in: 

• building a CI/CD pipeline using Jenkins, allowing us to automate and drastically shorten the process of app deployment and configuration;

• enabling several operators to work simultaneously during app deployment, instead of waiting for a single operator to build the Docker image;

• automation of files conversion process to speed up the app performance.

Final thoughts 

The customer’s existing infrastructure was unreliable with heavy workloads. IT Svit improvements automated and shortened the process of app deployment and configuration, as well as automated files conversion process that significantly sped up the app performance.

One of our UK customers suggested IT Svit build a blockchain-based customer loyalty program. We planed to use blockchain technology to issue discount coupons that would allow creating a mutually beneficial platform for both eCommerce shop owners and their customers. Additionally, this platform could not be hacked and data could not be lost or altered, and the origin of each discount coupon could be easily tracked. 

Case info

Location: UK

Partnership period: September-October 2017

Team size: 2 – 4 people

Team location: Kharkiv, Ukraine

Services: Blockchain development, cloud architecture, web development

Expertise delivered: AWS cloud administration, DevOps services, blockchain development, cloud infrastructure management

Technologies: Ethereum API, Laravel, Node.js.

Project requirements

The IT Svit team specified the following tasks:

• to design and develop a platform able to interact with Shopify and Ethereum APIs;

• to create a web portal for the retailers with simple and useful functionality for issuing the coupons;

• to develop a mechanism of simple gaining, transferring and spending discount coupons.

Challenges and Our solutions

Challenges: 

• platform should support online shops on Shopify, WooCommerce and Magento;

• shop owners can issue any amount of discounts and specify the rules of applying them;

• customers can use the coupons themselves anywhere across the network or give them as a gift to any other user;

• platform should allow tracking each coupon, providing analytics on where and how they are applied. This allows the shop owners to adjust their offers and increase the profits.

Our solutions: 

1. we created an intuitive and useful web portal for online retail owners and shoppers;

2. we developed an integrated platform working with Ethereum, Shopify, Magento and WooCommerce APIs;

3. we enabled the customers to save money and get discount coupons in any shop on the network.

Results

Blockchain technology ensures the safety of user’s data. The shop owners gain a powerful analytics tool, enabling them to track their coupons across the network. This allows adjusting their offers to maximize the user experience. Moreover, the number of cross-sales and up-sales grew by 75%, as the customers are incentivized to save money while making purchases.

Final thoughts 

IT Svit has developed an easy-to-use and efficient platform for issuing, transferring, applying and tracking discount coupons through a network of Shopify, WooCommerce and Magento eCommerce shops.

The IT Svit team decided to use blockchain technology to replace the excise stamps with QR-codes because this would provide benefits such as:

• reduction of excise stamps spendings by approx 90%;

• reduction of fraud risk to 0;

• increase the tax revenues by nearly 100%.

Case info

Location: Ukraine

Partnership period: September-October 2017

Team size: 2 – 4 people

Team location: Kharkiv, Ukraine

Services: Blockchain development, cloud architecture, web development

Expertise delivered: AWS cloud administration, DevOps services, blockchain development, cloud infrastructure management

Technologies: Ethereum API, Laravel, Node.js.

Project requirements

The QR-code should be used the same way an excise stamp is used nowadays, just without the possibility to forge it. It should be issued by the governing body and store the data about each excise good unit.

Challenges and Our solutions

Challenges:

• the system should be to highlight the status of each item, whether it is “packaged”, “delivered” or “sold”;

• the open ledger should ensure the records cannot be altered and the destination of each item can be tracked;

• the codes cannot be forged.

Our solutions:

• we usedEthereum smart contracts that allowed us to deliver a flexible and highly configurable platform that can interact with multiple third-party software modules through API;

• we configured MongoDB in the cloud to store the textual files and only the transaction hashes are stored within the blocks. This leads to minimal usage of system resources and allows the transactions to happen in milliseconds;

• we delivered both a web portal and a lightweight blockchain wallet for checking the QR-code status through a mobile device.

Results

Here is how our system works:

1. regulator prints the QR-codes;

2. manufacturer marks the goods with QR-codes and packs them;

3. tokens are put into the “packaged” state;

4. once the wholesale distributor splits the batch, the tokens are updated and the pallet codes are activated;

5. customers scan the item QR-codes to ensure their validity;

6. once the item is sold, its token is put to the “sold” state, closing the block.

Final thoughts 

Replacing the expensive (and forgeable) excise stamps with low-cost (yet unforgeable) QR-codes would ensure both cutting the stamp production costs and increasing the tax revenues from the distribution of excise goods.

Skillbyte GmbH with its custom Content Management System (CMS) contacted IT Svit with the request to build an out-of-the-box solution that uses cloud technology to rapidly deliver scalable, easily configurable and manageable infrastructure to showcase a web-based CMS.

Case info

Location: Cologne Area, Germany

Partnership period: May 2016 – ongoing

Team size: 2 – 8 people

Team location: Kharkiv, Ukraine

Services: Cloud infrastructure design and development, DevOps-as-a-Service, Big Data solutions, Web Development, QA and testing, product and infrastructure monitoring

Expertise delivered: AWS cloud administration, DevOps services, microservices and containerization, DevOps support team, cloud monitoring solutions

Technologies: AWS, Kubernetes, Docker, Ansible, Jenkins

Project requirements

The IT Svit team specified the following tasks:

• to migrate the showcase from Mesos/Marathon environment to Kubernetes;

• to shorten the product deployment time;

• to add the missing features;

• to configure error-proof update and recovery procedures;

• to introduce the Continuous Delivery pipeline to automate the setup and configuration;

• to split the monolith app into microservices to provide stability;

• to containerize the components for scalability;

• to deploy an in-depth monitoring and logging solution.

Challenges and Our solutions

Challenges:

• CMS itself is a monolith product built in Java, with more than 100 modules;

• more than 15 processes interacting with each other, creating a high workload;

• setup process is too complicated.

Our solutions:

• we performed a test deployment of the CMS and analyzed all the dependencies and possible bottlenecks;

• we moved the CMS from a Mesos/Marathon environment to a redundant Kubernetes cluster;

• we containerized the system components using Docker;

• we enabled the Continuous Implementation / Continuous Delivery (CI/CD) pipeline using Jenkins;

• we automated the product build and deployment using Ansible; 

• we automated the process of recovery for all the components in case of an interruption during the system update;

• we enabled the auto-scaling for the front-end service and automated its release cycle;

• we designed and developed a customized monitoring tool. 

Results

The process of the CMS setup and deployment is now fully-automated and the product has:

• 3 commands and around 15 minutes to deploy the CMS (previously it took 2 weeks to deploy);

• auto-scaling of the cluster nodes (to handle the load and optimize the costs);

• in-depth logging and analytics;

• real-time monitoring and smart alerts.

Final thoughts

The product is now enterprise-grade and production-ready. A truck manufacturing giant, several leading media, popular retailers and many other high-level EU enterprises have already appreciated this setup advantage.

Skillbyte GmbH with its custom Content Management System (CMS) contacted IT Svit with the request to build an out-of-the-box solution that uses cloud technology to rapidly deliver scalable, easily configurable and manageable infrastructure to showcase a web-based CMS.

Case info

Location: Cologne Area, Germany

Partnership period: May 2016 – ongoing

Team size: 2 – 8 people

Team location: Kharkiv, Ukraine

Services: Cloud infrastructure design and development, DevOps-as-a-Service, Big Data solutions, Web Development, QA and testing, product and infrastructure monitoring

Expertise delivered: AWS cloud administration, DevOps services, microservices and containerization, DevOps support team, cloud monitoring solutions

Technologies: AWS, Kubernetes, Docker, Ansible, Jenkins

Project requirements

The IT Svit team specified the following tasks:

• to migrate the showcase from Mesos/Marathon environment to Kubernetes;

• to shorten the product deployment time;

• to add the missing features;

• to configure error-proof update and recovery procedures;

• to introduce the Continuous Delivery pipeline to automate the setup and configuration;

• to split the monolith app into microservices to provide stability;

• to containerize the components for scalability;

• to deploy an in-depth monitoring and logging solution.

Challenges and Our solutions

Challenges:

• CMS itself is a monolith product built in Java, with more than 100 modules;

• more than 15 processes interacting with each other, creating a high workload;

• setup process is too complicated.

Our solutions:

• we performed a test deployment of the CMS and analyzed all the dependencies and possible bottlenecks;

• we moved the CMS from a Mesos/Marathon environment to a redundant Kubernetes cluster;

• we containerized the system components using Docker;

• we enabled the Continuous Implementation / Continuous Delivery (CI/CD) pipeline using Jenkins;

• we automated the product build and deployment using Ansible; 

• we automated the process of recovery for all the components in case of an interruption during the system update;

• we enabled the auto-scaling for the front-end service and automated its release cycle;

• we designed and developed a customized monitoring tool. 

Results

The process of the CMS setup and deployment is now fully-automated and the product has:

• 3 commands and around 15 minutes to deploy the CMS (previously it took 2 weeks to deploy);

• auto-scaling of the cluster nodes (to handle the load and optimize the costs);

• in-depth logging and analytics;

• real-time monitoring and smart alerts.

Final thoughts

The product is now enterprise-grade and production-ready. A truck manufacturing giant, several leading media, popular retailers and many other high-level EU enterprises have already appreciated this setup advantage.

One of the IT Svit customers turned to the IT Svit team with the request to build a user-friendly news aggregator platform from scratch. The client intended to have a platform where each customer could form a unique, personalized news feed, based on their preferences.

Case info

• Location: Netherlands

• Partnership period: December 2016 – July 2017

• Team size: 2 people

• Team location: Kharkiv, Ukraine

• Services: Big Data Development, Machine Learning, Web Development, Data Science, Cloud Architecture, Python Development

• Expertise delivered: Big Data solutions, training ML models, a rapid detection system for checking the news source updates

• Technologies: Python, Django, Docker, jQuery, Theano, aiohttp, MongoDB, SphinxSearch.

Project requirements

The IT Svit team distinguished the following tasks:

• to have the news data frequently updated;

• to have the news data categorized;

• to have the news data checked. For it to be credible, only primary sources should be given.

Challenges and Our solutions

Challenges:

• data set must be frequently updated and kept relevant;

• news must be categorized to simplify the navigation and personalize the user experience;

• only the primary source of the news must be listed.

Our solutions: 

• we built a web scraper tool utilizing the aiohttp library. The data gathered was stored in MongoDB;

• we created several Machine Learning models utilizing Theano and used these ML algorithms to filter out the irrelevant and/or unwanted content;

• we built the front-end using jQuery, and the back-end instances were containerized and automatically scaled using Docker;

• we used SphinxSearch to allow the customers to create their personalized news feeds;

• we created a huge base of news sources worldwide, split by industries, locations and spheres of interest;

• we built-in a highly performant tool that swiftly analyzed the article content to check if it was updated.

Results

IT Svit has built a powerful, flexible and highly performant news aggregation platform. This news base is rapidly updated, using a highly performant tool. The platform can create a personalized news feed for any customer, based on their preferences and settings. Moreover, the aggregator is easily embedded into any website as a plugin with a simple configuration.

Final thoughts 

The IT Svit team created a platform for automated collection, verification, deduplication and classification of industry-specific news from a variety of credible sources. These news feeds can be both shown on a web portal, and/or can be embedded into any website.

IT Svit customers mentioned a lot that they pretty often face one and the same problem - they cannot find some piece of information they need on their site fast. That is when the IT Svit team decided to create a bespoke web scraper tool that can be added to a particular website and will create custom search indexes.

Case info

Location: Kharkiv, Ukraine

Partnership period: August 2015 – February 2018

Team size: 2 – 4 people

Team location: Kharkiv, Ukraine

Services: Cloud infrastructure design and development, Python development, Data Science, Big Data solutions, Machine Learning algorithms

Expertise delivered: Cloud infrastructure design and implementation, Python development, Big Data architecture design and management

Technologies: Python, asyncio, aiohttp.

Project requirements

The IT Svit team specified the following tasks: 

• to create lightweight, simple, yet efficient web crawlers;

• to design a user-friendly interface;

• to develop tools with low hardware requirements.

Challenges and Our solutions

Challenges:

We were supposed to built-in search tools that have:

• high performance;

• low system resource consumption;

• ease of configuration;

• ease of use.

Our solutions:

We used asyncio and aiohttp Python libs to create a tool that:

• comes with a built-in webserver to ensure the simplicity of launching it;

• can be easily integrated into any website;

• can be utilized through any browser;

• has low hardware requirements;

• works quickly.

Results and Final thoughts 

IT Svit developed the required web scrapers and other Big Data solutions to enable our customers to form the data set for training their search engine. The main advantages of our searching tool, precisely its usability and speed were appreciated by our clients. Toweya, let us say, provided us basic specifications and we helped them create an easy to use and performant search engine platform, which enables incremental web search and provides accurate results.

The web and mobile app developer Everdapt contacted IT Svit with the request to provide the AWS infrastructure for their web app that uses the Neo4j database. The problem was that Neo4j does not have native scalability and redundancy features, so this functionality had to be implemented.

Case info 

Location: France

Partnership period: 3 weeks in December 2017

Team size: 2 people

Team location: Kharkiv, Ukraine

Services: Cloud infrastructure design and development, DevOps-as-a-Service, Web Development, product and infrastructure monitoring

Expertise delivered: AWS cloud administration, DevOps services, microservices and containerization, DevOps support team, cloud monitoring solutions

Technologies: AWS CloudFront CDN, Neo4j DB, Amazon Load balancer, Amazon EC2 and EBS, Kubernetes, Docker, Ansible, Jenkins.

 Project requirements

The IT Svit team decided to:

• design the cloud architecture for deploying a web solution, using a specific database – Neo4j;

• ensure the system scalability and fault-tolerance;

• ensure the system security and redundancy;

• prepare configuration and deployment scripts and guidelines.

Challenges and Our solutions 

Challenges: 

• redundancy for the Neo4j database;

• scalability, security and usability of the system.

Our solutions 

The app consisted of three components: Customer-facing UI, API, and a Neo4j database. With that in mind, we decided to split the application deployment into three parts:

• Amazon S3 storage behind the Amazon CloudFront CDN for the Frontend part;

• several API instances in different Availability Zones behind the Amazon Application Load Balancer (ALB);

• Amazon Elastic Compute Cloud (EC2) instance for Neo4j database engine with Amazon Elastic Block Store (EBS) volume attached for data persistence.

Both the UI and the API parts operate in private networks, protected by Amazon CDN and Amazon ALB. This way the origin instances where the application is running are not available directly from the Internet. A bastion host is used for management tasks.

Results

1. We used AWS CloudFront CDN for the front-end to ensure the system’s scalability and AutoScaling groups for the back-end; 

2. A bastion host is ensuring access security;

3. The application runs on several nodes located in various Amazon availability zones behind an Amazon Load Balancer to ensure optimal performance

4. The system is modular. It means the component is rebooted automatically, if it fails. 

Final thoughts 

The IT Svit DevOps team analyzed the client’s existing system configuration and came up with recommendations on ensuring its stable performance. The solutions were built-in scalability and redundancy features for Neo4j that were provided externally.

IT Svit decided to implement internal monitoring for our services with Zabbix-server, Prometheus operator and Grafana components.

Case info

Location: Ukraine

Partnership period: 2005-ongoing

Team size: 2 – 4 people

Team location: Kharkiv, Ukraine

Services: Cloud architecture, cloud infrastructure management, cloud monitoring solutions

Expertise delivered: AWS cloud administration, DevOps services, cloud infrastructure management, monitoring solutions configuration

Technologies: Zabbix, Kubernetes, Prometheus, Grafana.

Project requirements

We wanted our system to provide:

• automatic monitoring of various cloud infrastructure parameters like CPU usage, bandwidth usage, disk volume usage, etc;

• timely alerts if any issues occur;

• informative notifications with screenshots.

Challenges and Our solutions

Challenges: 

• automatic monitoring performed by Zabbix;

• timely alerts sent to Telegram;

• informative screenshots attached.

Our solutions:

• Zabbix + Prometheus as a data source for Grafana. This allowed us to enable detailed and on-point issue-reporting;

• Zabbix agents are running in every container or instance within our infrastructure and they report to a Zabbix server should any issue arise;

• Python script sends all the information about the incident to Telegram chat, so the admins are notified immediately.

Results

With Zabbix+Prometheus+Grafana monitoring installed, we always receive alert notifications in Telegram and know when something is wrong with our Linux server or Kubernetes cluster. It allows us to respond instantly and solve issues very fast.   Informative screenshots, when attached to the notifications, give us detailed information about the issue. As a result, we save much of our time dealing with bugs, because we know exactly what happened. 

Final thoughts 

The IT Svit DevOps team implemented an internal monitoring system to be better informed of the various processes within our IT infrastructure and to be able to identify and solve the issues faster. Combining Zabbix + Prometheus + Grafana allows us to monitor the IT infrastructure cost-effectively and almost effortlessly while immediately responding to the issues.

IT Service provider INTEX-consulting turned to the IT Svit team for us to develop a CMS solution with Laravel and Node.js.

Case info

Location: Russian Federation

Partnership period: October 2015 — April 2016

Team size: 2-6 people

Team location: Kharkiv, Ukraine

Services: Cloud infrastructure design and development, Full Stack Web Development, Laravel & Node.js development, QA and testing

Expertise delivered: AWS cloud administration, DevOps services, Laravel & Node.js development, QA and testing

Technologies: AWS, Laravel & Node.js, Critical CSS.

Project requirements

The IT Svit team specified the following project tasks: 

• to build a user-friendly and intuitive interface for the CMS;

• to endow CMS with all the basic functionality and advanced features built with the help of Critical CSS;

• to write backend using Laravel, the most popular PHP framework;

• to meet a 6-month deadline.

Challenges and Our solutions

Challenges:

• UI/UX development and customization;

• frontend functionality development using Critical CSS and Node.js;

• backend functionality development using Laravel;

• product testing and QA;

• strict deadline of 6 months.

Our solutions:

We introduced various custom modules:

• HTML-rendering with the DB templates using the dbview function allowed us to redefine the Laravel methods responsible for template discovery and configuration;

• RabbitMQ message order controller allowed us to manage the messages sent to RabbitMQ for better system stability and performance;

• customized request validation ensured some additional json processing used in request handling;

• critical CSS processing — a specific controller for generating dynamic pages using Critical CSS.

Results

The fact that the IT Svit team has polished internal workflows and on-point communication all the tasks were accomplished in time, within 6 months and the CMS, developed by IT Svit, works fast adding new items to websites with very little effort involved.

Final thoughts

IT Svit team was able to execute all the tasks in time, and the customer was delighted with our performance. Apropos, the product was sold more than 1,000 copies since 2016.

A stock-breeding concern specializing in breeding pigs contacted IT Svit with the request to develop a Big Data solution for data processing and visualization. 

Case info 

• Location: The United Kingdom

• Partnership period: September – December 2017

• Team size: 2 people

• Team location: Kharkiv, Ukraine

• Services: Cloud infrastructure design and development, Big Data solutions, Web Development, Python development, data visualization, data normalization

• Expertise delivered: AWS cloud administration, Machine Learning, DevOps services, Big Data architecture, Python development, data science, data visualization

• Technologies: PostgreSQL, ML in AWS, Python, Flask, Jupyter Notebook, Chart.js.

Project requirements

The customer wanted us to provide the following services:

• to develop a Big Data solution for processing data;

• to decrease the time needed to assess the data;

• to configure a user-friendly dashboard enabling the users to visualize the data.

Client’s challenges and Our solutions

Challenges: 

• existing data set improvement;

• Machine Learning model to track the effectiveness of the meds creating and implementing;

• dashboard for data visualization designing.

Our solutions:

• we wrote a data normalization tool in Python to ensure the uniformity of data;

• we used Python and JupyteR Notebook to makestatistical analysis and highlight patterns;

• cloud infrastructure was deployed in AWS and we used the AWS Machine Learning service in the process of training the model;

• data was stored in the cloud using the PostgreSQL database and Amazon RDS;

• we used Flask and Chart.js to enable the data visualization and created a flexible dashboard for viewing the graphs. 

Results

With ML algorithm based on the historical data implementation our client’s system has experienced the following improvements:

• - quick and precise diagnosis of the disease based on the symptoms;

• - recommendations for treatment are given;

• - the time between identifying the disease and applying the treatment,

• is significantly reduced.

 Final thoughts

The IT Svit team developed a solution for data processing and data visualization that increased our client’s business profitability and reduced the losses due to livestock diseases.

This solution consists of the CCTV cameras installed at the key roads of the city, paired with Raspberry Pi 2 or 3, and 3 or 4G modems. The cameras take photos of each passing car, the IoT node identifies the brand and the model, and this information is sent to the cloud processing server. The customer wanted to provide their users with clean and user-friendly tools for real-time analysis of the car brands and models currently in circulation in any city and they turned to the IT Svit team for us to help with designing and developing such a solution.

Case info 

• Location: Vienna, Austria

• Partnership period: May 2018

• Team size: 3 people

• Team location: Kharkiv, Ukraine

• Services: Cloud infrastructure design and development, Big Data solutions, Web Development, Python development, data visualization, data normalization, IoT

• Expertise delivered: AWS cloud administration, DevOps services, Big Data architecture, Python development, data science, data visualization, IoT architecture

• Technologies: Cassandra, Python, Django, Chart.js.

Project requirements

The IT Svit team distinguished the following tasks:

• IoT is to take the car photos, identify the model using OCR (Optical Character Recognition), send the report to the cloud server and delete the photo afterward to comply with GDPR;

• The back-end is to accept the reports, filter, aggregate and store the data in a  NoSQL database;

• The front-end is to represent the charts and reports and provide the end-user APIs to interact with any third-party analytical system.

Challenges and Our solutions

• Challenges: 

• - to design and develop the structure of an IoT node to be deployed on the streets;

• - to ensure the secure and reliable aggregation of the data in the cloud;

• - to provide the tools for data visualization;

• - to give the customers access to the historical data; 

• - to provide a clean API to enable interaction with third-party systems.

• Our solutions:

• - OCR algorithm identifies car brands and models seen in the photos;

• - photos are deleted on the spot when complied with the GDPR;

• - car brand and model are sent to the central server in AWS or GCP cloud;

• - data processed and stored in a Cassandra NoSQL database in the back-end;

• - customers interact via a web portal, where they can view the charts and reports, or get the data directly to their third-party software via API.

Results

Here is how our solution works:

1. CCTV cameras are installed at the key roads of the city, paired with Raspberry Pi 2 or 3, and 3 or 4G modems;

2. The cameras take photos of each passing car, the IoT node identifies the brand and the model, and this information is sent to the cloud-based processing server;

3. A dataset of car brands and models is formed in real-time. This data can be leveraged to deliver the predictive analysis for the market.

Final thoughts 

The system the IT Svit team developed has shown good results during the alpha-testing phase. One IoT node can identify about 150 cars per minute (a rush hour on a three-lane road), which makes the system quite feasible from the analytical perspective.

The Network Cloud Company needed the IT Svit DevOps team to upgrade their cloud infrastructure and automate their telecom product provisioning and deployment. They wanted to redesign the existing system to improve product stability and performance.

Case info 

• Location: Atlanta, Georgia

• Partnership period: January 2018 – ongoing

• Team size: 2 – 8 people

• Team location: Kharkiv, Ukraine

• Services: Cloud infrastructure design and development, DevOps-as-a-Service, product and infrastructure monitoring

• Expertise delivered: AWS cloud administration, DevOps services, DevOps support team, cloud monitoring solutions

• Technologies: AWS, Kubernetes, Docker, Ansible, Jenkins, Python CentOS 7.

Project requirements

IT Svit team defined the following tasks:

• to design & implement the cost-effective, performant and secure infrastructure for VoIP communication;

• to automate the provisioning of the backend infrastructure services;

• to design an architecture for a CRM to automate the customer provisioning;

• to design & implement the internal system for managing multiple third-party modules.

Challenges and Our solutions 

Challenges:

• automation tool to streamline the provisioning backend infrastructure deployment development;

• integration with a new CRM platform to automate the customer provisioning;

• the CRM designing;

• messaging platform designing

• backend functionality development to automatically integration new code.

Our solutions: 

• to audit the existing IT infrastructure;

• to automate cloud infrastructure;

• to engineer CRM platform;

• to design and implement the backend infrastructure services;

• to design and develop a messaging platform to automate the product updates;

• to integrate the updates from the third-party modules using the platform;

• to provide the ongoing cloud infrastructure management and monitoring.

Results

The IT Svit DevOps team’s work resulted in:

• product provisioning time shortened from 1,5 weeks to 1 hour;

• provisioning is provided 200-300 times a month, instead of 5-6;

• 30-35% of the budget is saved.

Final thoughts 

The IT Svit DevOps team upgraded and redesigned the cloud infrastructure for the Network Cloud Company and significantly improved their telecom product stability and performance.

The marketing and advertising agency needed a full-scale team of developers to build the frontend and backend for a trip-planning portal and a travel website and deliver a cloud infrastructure for these projects.

Case info

• Location: United States

• Partnership period: October 2016 — ongoing

• Team size: 2-8 people

• Team location: Kharkiv, Ukraine

• Services: Cloud infrastructure design and development, Full Stack Web Development, Python development, QA and testing

• Expertise delivered: AWS cloud administration, DevOps services, Python development, QA and testing

• Technologies: AWS, Python, Django, AngularJS.

Project requirements

The customer needed to create two websites — a travel website and a trip planning portal. IT Svit team specified the following tasks:

• to write the back-end for two websites using Python and Django;

• to write the frontend part with AngularJs;

• to provide the cloud infrastructure to support the development and functioning of these websites;

• to provide the QA and testing of the functionality.

Challenges and Our solutions 

• Challenges:

•  - the portal had to be able to interact with other travel industry tools through  an API;

•  - the travel website had to allow the customers to provide feedback while  remaining anonymous;

•  - IT Svit had to provide the marketing support, content creation and technical documentation for these websites.

• Our solutions:

•   - we built a regular reporting process to keep the PM overseas up-to-date with the development progress;

•  - we developed a clean API to allow the platforms to interact with third-party modules;

•  - we provided the full cycle of infrastructure provisioning services, from delivering the testing environments and building CI/CD pipelines, to managing the ongoing maintenance of the websites.

Project results

This project required the full cycle of software development and IT Svit delivered: 

• full-stack development for both websites;

• cloud infrastructure design for their deployment;

• ongoing QA and improvement of the services.

Final thoughts 

The talented IT Svit DevOps team has built frontend and backend for a trip-planning portal and a travel website and designed a cloud infrastructure for marketing and advertising agency that allowed our customer to gain new clients and provide better services.

IT service management provider Preterion contacted IT Svit with the request for our DevOps team to assist with a wide range of tasks and augment their internal team with our skillset.

Case Info 

Location: Kellyville Ridge, NSW, Australia

Partnership period: January 2017 – ongoing

Team size: 2 – 8 people

Team location: Kharkiv, Ukraine

Services: Cloud infrastructure design and development, DevOps-as-a-Service, Web Development, QA and testing, product and infrastructure monitoring

Expertise delivered: AWS cloud administration, DevOps services, microservices and containerization, Perl development, PHP development, DevOps support team, cloud monitoring solutions, Big Data solutions

Technologies: AWS, Kubernetes, Docker, Ansible, Jenkins, PHP, Perl, MySQL, Linux.

Project requirements

Preterion needed the IT Svit DevOps team to provide such services:

• Linux-based DevOps software development

• PHP-based CRM development

• web-based app in Perl

• Various fixes and improvements of existing cloud monitoring solutions

• Ongoing cloud monitoring and support

• Enabling CI/CD pipelines

• IT Infrastructure management automation

IT Svit DevOps team specified the following tasks:

• to design and implement various cloud infrastructures;

• to fix bugs for multiple apps;

• to develop and deploy CRM;

• to manage and optimize multiple IT infrastructure tasks (implementing xWiki backups, Your Backup tool, implementing Baruwa Spam Protection solution, etc.)

Results 

Our work resulted in:

• in-depth analysis of IT infrastructures to identify possible bottlenecks;

• proactive consulting on system improvements;

• implementation of improvements;

• automation of infrastructure deployment & management through scripts;

• providing the detailed developer documentation for each module built;

• training the customer’s team to leverage the system best.

Final thoughts

IT Svit has handled all the tasks successfully and currently provides ongoing support for various Preterion projects through a ticketing system. It Svit DevOps team is currently involved in:

• team cloud architecture delivery and management

• web development and QA

• various software maintenance and updates.

This allowed Preterion to engage with various projects as well as decrease the ticket processing time by 50% and average operating expenses by 25-30%.

Preterion, an Australian cloud migration service provider, turned to IT Svit for us to assist with a wide range of tasks and augment their internal team with our skillset. 

Case Info

• Location: Kellyville Ridge, NSW, Australia

• Industry: Cloud management, IT Security, DevOps services

• Partnership period: January 2017 – ongoing

• Team size: 2 – 8 people

• Team location: Kharkiv, Ukraine

• Services: Cloud infrastructure design and development, DevOps-as-a-Service, Web Development, QA and testing, product and infrastructure monitoring

• Expertise delivered: AWS cloud administration, DevOps services, microservices and containerization, Perl development, PHP development, DevOps support team, cloud monitoring solutions, Big Data solutions

• Technology stack: AWS, Kubernetes, Docker, Ansible, Jenkins, PHP, Perl, MySQL, Linux.

Project Requirements 

Project requirements included: 

• Linux-based DevOps software development;

• PHP-based CRM development;

• web-based app in Perl development;

• cloud monitoring and support providing;

• CI/CD pipelines enabling;

• IT Infrastructure management automation.

Challenges and Our solutions

Challenges: 

• design and implementation of various cloud infrastructures;

• bug fixes and deployment automation for multiple apps;

• CRM development and deployment;

• IT infrastructure management and optimization tasks (implementing xWiki backups, Your Backup tool, implementing Baruwa Spam Protection solution, etc.)

Our solutions: 

• analysis of IT infrastructures to identify possible bottlenecks;

• proactive consulting on system improvements;

• implementation of improvements;

• automation of infrastructure deployment & management through scripts;

• providing the detailed developer documentation for each module built;

• training the customer’s team to leverage the system best.

Results

IT Svit has handled all the tasks successfully and currently provides ongoing support for various Preterion projects through a ticketing system. IT Svit DevOps team is engaged at:

• cloud architecture delivery and management;

• web development and QA;

• various software maintenance and updates. 

Our DevOps services allowed Preterion to engage with the projects outside the scope of their abilities and reap the benefit of:

• significant skillset extension;

• reduction of technical support ticket response time by nearly 50%;

• IT infrastructure management optimization and expenditure cuts by 25-30%.

Feedback

Here we present feedback from our customer instead of conclusions. Cristian Iconomu, Director at Preterion Pty Ltd, said about IT Svit,

“Given their successful track record and wide skill set, IT Svit now handles all our tasks internally, from full-scale projects to complex technical portions. Their integrity, responsiveness, and compelling price point continue to promote a productive partnership.”

10 years of expertise in software development have helped the IT Svit team build Hurma System — an innovative all-in-one solution for managing HR processes, recruitment and OKR. The IT Svit DevOps team is supposed to enable seamless software development and infrastructure management operations in production.

Case info

Project: Hurma System

Location: Kharkiv, UA

Industry: HRM/HRIS development 

Partnership period: November 2017 — ongoing

Team size: 30

DevOps Team: 1 DevOps Architect, 2 DevOps Engineers

Team location: Kharkiv, Ukraine

Services: Cloud infrastructure design and optimization, CI/CD implementation, app containerization, API integration, monitoring implementation

Expertise delivered: App containerization and container management, cloud infrastructure management, monitoring implementation, CI/CD configuration, API integration

Technology stack: PHP/Laravel, Python/Django, MariaDB, Google API, Terraform, Docker, Kubernetes, Helm, Nginx Ingress controller, Ansible, Jenkins Pipeline, Prometheus, Grafana, Sentry.

Project requirements

IT Svit DevOps team worked as an insourcing contractor for the Hurma System development team that needed:

• to construct infrastructure for product development from scratch; 

• to configure CI/CD pipelines to automate code testing and deployment;

• to implement in-depth monitoring to track the system performance in production.

Challenges and Our solutions

Challenge: during the MVP development, the deployment was manual and took 3 hours that was clearly too long for the needs of the production environment. 

Our solution: CI/CD pipelines implementation. To do this, IT Svit system architects prepared a unified Docker Image of the product and updated the Kubernetes manifest templates. Jenkins Pipeline was used to automatically deploy the new product versions, thus shortening the update time to 30-40 minutes. Hurma System is actively developed. With that in mind, it was needed to speed up the update process even more. IT Svit DevOps specialists leveraged the seed migrations mechanism in Jenkins Pipeline, which allowed them to simplify and speed up the update process to 2 minutes. There is more to come! When the team started updating 200+ environments, the manual configuration of variables for each instance was no longer feasible. DevOps engineers went with the Infrastructure as Code approach and created an additional Jenkins pipeline that managed Jenkins pipelines for individual customer instances and ingested the required parameters from the VCS. Using Helm to create and manage templates for Kubernetes operations helped automate most infrastructure management processes, so now IT Svit can deploy a fresh instance of Hurma System in 5 minutes and update it to a new version in around 1 minute.

Results 

Our work on the Hurma system has resulted in:

• more than 400 Single Tenant environments containing the system instance running in the cloud;

• individual environment configuration time shortened from 3 hours to 2 minutes;

• time-to-market for new features - 2-3 weeks tops.

Final thoughts

Extensive IT Svit DevOps expertise helped build a powerful and cost-efficient software development process for Hurma System, as well as enabling in-depth monitoring of the Kubernetes cluster in production using Prometheus, Grafana and Sentry.

IT Svit received a request to assist with moving an enterprise SIEM (security information event management) system from Rancher to a Kubernetes deployed at Tectonic, a commercial Kubernetes+CoreOS platform. 

Case info

Location: New York, USA

Industry: Enterprise cybersecurity

Partnership period: May 2019 – ongoing

Team size: 1 Team Lead, 1 DevOps engineer

Team location: Kharkiv, Ukraine

Services: App containerization, cloud transition, CI/CD and monitoring implementation

Expertise delivered: App containerization, cloud infrastructure management and optimization, monitoring and alerting implementation, CI/CD configuration

Technology stack: Tectonic, Kubernetes, AWS, DigitalOcean, Azure, Google Cloud, Docker, OVA, VirtualBox, VMWare, HyperV.

Project requirements

The IT Svit team decided to take two main steps:

Create Docker images and docker files for all system components:

• ELK stack with SearchGuard;

• front-end web app for interacting with all components;

• RabbitMQ message broker;

• Grafana/Prometheus for monitoring;

• MineMeld for threat intel aggregation;

• dradis for threat assessment;

• Opmantek for a security audit.

Design high-performance environments for various hosting platforms with the following features:

• high-availability;

• failover;

• fast response times;

• optimized for deployment;

• stateful deployment for Elasticsearch data nodes;

• CI/CD deployment pipelines;

• autoscaling;

• monitoring tools.

Challenges and Our solutions

With the customer providing detailed all interactions, IT Svit DevOps engineers quickly elaborated scenarios for automated deployment and configuration of a Kubernetes cluster to various hosting platforms utilizing our extensive expertise with AWS, Google Cloud and DigitalOcean. We also provided the tools for monitoring the system in production and detailed developer documentation on platform installation and configuration.

Results

The customer received scenarios for the error-free automated deployment of their platform to a variety of hosting providers that allowed them to increase their target audience and lower the learning curve.

Final thoughts

IT Svit DevOps team has shown an in-depth understanding of app containerization best practices and configuration of scalable, secure, highly-performant and cost-efficient environments on various cloud platforms. The result of our work not only met but exceeded our customer’s expectations.

German Airbnb online marketplace MICE contacted IT Svit and wanted us to propose them more flexible and durable hosting and implement CI/CD pipelines to enable seamless updates to multiple Ruby-on-Rails apps.

Case info

Location: Attenkirchen, Germany

Industry: Venue booking for events and meetings

Partnership period: January 2019 – December 2019

Team size: 1 Team Lead, 2 DevOps engineers

Team location: Kharkiv, Ukraine

Services: AWS infrastructure performance optimization, CI/CD and monitoring implementation

Expertise delivered: Cloud infrastructure management and optimization, monitoring and alerting implementation, CI/CD configuration

Technology stack: AWS, Ruby on Rails, Redis, ElasticSearch, PostgreSQL, Memcached, Kubernetes, Docker, Jenkins, Helm, ELK stack, elastalert, Prometheus, Grafana.

Project requirements

The project plan included:

• migrating the existing containerized microservices with Ruby on Rails apps to a new cloud platform (AWS) with the help of Helm;

• implementing the CI/CD pipeline using Jenkins;

• building a unified monitoring platform based on Prometheus, Nagios and ELK stack;

• preparing detailed developer documentation and share info with the MICE team.

Challenges and Our solutions

Clear and quick communication with MICE’s CTO helped us a lot to follow a predefined roadmap for the project and pinpoint possible roadblocks. Some additional challenges arose that were outside the initial scope of work, but the Agile approach of the IT Svit team allowed to include additional tasks to the project pipeline on the go.

Results

MICE’s online platform got:

• new cloud hosting set up on AWS which helped the platform to sustain heavy traffic loads;

• automated CI/CD pipelines enabled development process escalation;

• effective monitoring system;

• documentation to help the Mice team master the new system quickly.

Final thoughts

Our DevOps team gained amazing experience with complex microservice systems and the MICE platform got more flexible and durable hosting and CI/CD pipelines that enable seamless updates to multiple Ruby-on-Rails apps that allow it to serve 30% more customers and reduce timeframes for new releases.

Anthems, a media portal from the USA, required our help with infrastructure optimization. More specifically Anthems wanted us to audit existing infrastructure and suggest ways to ensure that the platform will be able to handle millions of users simultaneously.

Case info

Location: The United States

Industry: Social Media Network

Partnership period: June 2019 – ongoing

Team size: 2-3 DevOps specialists

Team location: Kharkiv, Ukraine

Services: Cloud infrastructure design and implementation, database optimization, Web Development, load-testing

Expertise delivered: GCP cloud administration, DigitalOcean cloud, DevOps services, PHP app architecture, SQL and NoSQL database management, cloud monitoring 

Technology stack: GCP services, JMeter, DO, Ansible, MySQL, NoSQL, PHP.

Project requirements

Anthems required our DevOps team to:

• perform a load-testing to ensure that the platform can handle millions of users;

• test the current API to see how many users they can handle;

• leverage monitoring tools to identify if any scalability bottlenecks exist due to poor PHP code VS poor MySQL queries VS poor MySQL index VS suboptimal server configuration;

• design and implement the solutions for the problems found;

• conceptualize other ways to increase scalability and performance;

• design and develop infrastructure using DevOps best-practices.

Challenges and Our solutions

The main challenges are:

• data and content stored in MySQL (not the fastest DB);

• resource-intensive queries in the common DB.

Our solutions:

• suggested introducing separate tables for every user;

• suggested storing all user data in some NoSQL database and store only the link to it in the common MySQL database. 

Results

• IT Svit DevOps team: 

1. enabled and configured correct Google CDN caching, which allowed the system to process 2,000 rps with 250 ms response time under 90% percentile;

2. wrote a detailed description of the infrastructure and workflows in place;

3. created an Ansible playbook for automated creation of load-testing infrastructure on DigitalOcean;

4. selected the safest cache for the system;

5. started logging the SQL query caching to ensure it helped speed up the system operations;

6. configured system monitoring and alerting using Google Stackdriver;

7. ensured the system security by hiding the instances behind Cloudflare CDN into a private subnet behind a bastion host to avoid direct access via the Internet;

8. created auto-scaling groups for API and PHP endpoints to ensure simple scalability under heavy workloads.

Final thoughts

The IT Svit devoted team helped Anthems with infrastructure optimization and continues working on Anthems’ PHP and database architecture improvement because we can always do better.

The cryptocurrency exchange allowing to instantly trade in various cryptocurrencies including Ethereum turned to the IT Svit DevOps team with the request to optimize the resource usage of the existing cluster, implement automated MongoDB backups and enable smart monitoring of the key system parameters.

Case info

Location: London, UK

Industry: Cryptocurrency trading

Partnership period: October 2019 – ongoing

Team size: 1 Team Lead, 1 DevOps engineer

Team location: Kharkiv, Ukraine

Services: MongoDB database performance optimization, monitoring implementation, automation of backup/restore operations

Expertise delivered: Cloud infrastructure management and optimization, monitoring implementation, CI/CD configuration

Technology stack: AWS, GitHub, Ansible, Kubernetes, Helm, Terraform, MongoDB, Prometheus, Grafana, AlertManager.

Project requirements

The customer wanted our DevOps team to:

• optimize Prometheus+Grafana monitoring to identify slow MongoDB queries;

• optimize Kubernetes cluster performance;

• automate backup/restore operations for MongoDB;

• enable archiving and retrieval for historical MongoDB data;

• build a staging environment that runs in parallel;

• minimize the expenses of the system operations.

Challenges and Our solutions

The main challenge is that  MongoDB had some long queries that used an exorbitant amount of resources. Additionally, Prometheus and Grafana versions were outdated and not configured quite correctly. Finally, he lacked an inexpensive staging environment to test the system changes before rolling them out to production.

We solved this in three steps:

1. configured the backup and restore for MongoDB. Luckily, IT Svit has developed our solution for automated MongoDB backup/restoration in AWS using Terraform and Jenkins;

2. optimized the MongoDB performance. We used the database sharding. Another solution was to use specific node pools for specific workloads (one for monitoring, one for staging, one for prod, one for MongoDB). We have also configured limits to resource requests. This helped reduce resource usage and lower costs;

3. improved the efficiency of cluster monitoring. We utilized Helm charts to build and update Prometheus and Grafana operators to the latest versions and configure their interaction with AlertManager. 

Results

Our customer currently has a system running on a Kubernetes cluster with automated backup and restoration capabilities. By sharding the database we helped lower the system workload and decrease the expenses. 

Final thoughts

The IT Svit DevOps team helped lower our customer’s resource consumption and operating costs significantly and is currently working on further improving the system performance.

A cryptocurrency platform operating the most secure and cost-effective block production nodes for decentralized PoS protocols contacted the IT Svit DevOps team with the request to restructure the system layout and add new cryptocurrencies to the range of offers.

Case info

Location: USA

Industry: Cryptocurrency stacking and lending

Partnership period: August 2019 – ongoing

Team size: 1 Team Lead, 2 DevOps engineers

Team location: Kharkiv, Ukraine

Services: AWS infrastructure performance optimization, CI/CD and monitoring implementation

Expertise delivered: Cloud infrastructure management and optimization, monitoring and alerting implementation, CI/CD configuration

Technology stack: AWS tools, Kubernetes, Docker, Ansible, Jenkins, Prometheus, Grafana, AlertManager, MySQL Aurora, SMTP2Go, SumoLogic, OpenVPN, EFS Provisioner, Ingress controllers, Certificate manager, Falco, Vault.

Project requirements

Our customer needed help with:

• building a multi-cloud architecture for running containerized blockchain nodes;

• containerizing all cryptocurrency apps running in the system;

• configuring a Kubernetes cluster to run EOS and deploy all the pods/services/deployments prepared;

• ensuring Multi-Availability-Zone deployments within a single AWS region enabling minute monitoring of system performance.

Challenges and Our solutions

Here are the challenges and our solutions. So, it’d be great if we could ensure:

• fault tolerance with minimal downtime for all blockchains supported. Our solution — We deployed the system in 2 regions with the same infrastructure and launched a HashiCorp Consul server tracking the state of the active region. 

• automated system failover and data recovery. Our solution — we enabled separate pod backups for every region and created a script checking the timestamp of the last file update.

• continuity of operations and automated bans for blockchains after 30 minutes out of consensus. Our solution — all logs are monitored with Log exporter, alerts are sent to Alert Manager in Prometheus.

• automated checks for new blockchain versions. Our solution — We configured 2 Jenkins jobs. The first job checks for the new blockchain app version, the second one builds a new image and sends a notification to Slack.

• real-time two-level monitoring. Our solution — We use Falco and custom rules to monitor external container access attempts or incorrect node behavior.

• robust user authentication and access rights configuration.
  Our solution — We used kiam (Kubernetes Identity & Access Management) agents to configure limited access rights assigned to each pod.

Results

IT Svit team was able to configure the multi-cloud architecture able to run on AWS resources, IBM Cloud and Google Cloud Platform that allows us to run Algorand,  Cosmos, Decred and many other blockchains. Infrastructure as Code approach has empowered the speed of the new blockchains onboarding. Two-level real-time monitoring saves a lot of time and effort for tracing and issues down. As a result, the system elastically scales up to handle heavy workloads.

Final thoughts

The IT Svit team was able to provide all the deliverables on time as once the first set of blockchain apps was containerized and the monitoring implemented, the only remaining challenge was to add new cryptocurrency to the system as quickly as it was available.

An outsourcing company IT Svit, one of the most trustful Managed and DevOps services providers, developed Hurma System as a SaaS solution for recruitment and HR management. This time our client is our own company and the requirement was to deliver new video features for Hurma. Speaking of video call features, IT Svit is good at WebRTC implementation and has published the source code for its online conferencing tool prototype — https://github.com/it-svit/ownconf.

Case info

Location: Kharkiv, Ukraine

Industry: Recruiting, human resources

Partnership period: Dec 2019 – ongoing

Team size: 1 Python developer, 1 DevOps engineer, 1 Web developer

Team location: Kharkiv, Ukraine

Services: Online conferences subsystem development, application containerization, CI/CD and monitoring implementation

Expertise delivered: Requirements analysis, integration with the existing system, performance optimization

Technology stack: WebRTC, Python 3, Docker, Coturn, Vue.js, ffmpeg, AWS S3

Project requirements

The customer provided project specifications. Online video conferencing feature should:

• work directly from the browser without any additional software or extensions;

• support mobile devices, as well as multiple desktop platforms (Windows/Mac/Linux);

• should include screen sharing, chat, audio muting, speaker highlighting;

• provide automatic recovery on disconnection and easily switch between WiFi and 3G/4G/5G/LTE;

• have a user-friendly interface.

Challenges and Our solutions

The main challenge is to build a platform with a stable Internet connection. Our solution is to develop a specific component for client applications enabling them to switch to alternative Internet access routes with minimum connection disruption. A built-in WebRTC feature constantly monitors the connection performance and lowers the bitrate of video/audio streams automatically if the connection becomes worse.

Results

Here are our results: 

• video calls are integrated into the Hurma system;

• video calls can be performed directly from the browser, any browser;

• a new feature to record the calls is added.

Final thoughts

The peer-to-peer approach allowed us to minimize resource consumption and the simple API of the video conferencing system allowed us to customize UI easily. The talented IT Svit DevOps team provided a simple and reliable tool for online communications directly from any browser.

An influencer marketing cloud for enterprise creators that offers private enterprise cloud services CreatorIQ turned to IT Svit with the request of infrastructure optimization and migrating the legacy infrastructure to AWS.

Case info

Location: Los Angeles, CA

Industry: Private enterprise cloud 

Partnership period: July 2019 — January 2020

Team size: 1 team lead, 2 DevOps engineers

Team location: Kharkiv, Ukraine

Services: Cloud infrastructure design and optimization, database performance optimization, monitoring implementation, CI/CD implementation, app containerization, API integration

Expertise delivered: App containerization and container management, cloud infrastructure management, monitoring implementation, CI/CD configuration, API integration

Technology stack: AWS, GitHub, Jenkins, MS SQL, AWS Lambda and Redshift, ElasticSearch, Node.JS, social media APIs (Youtube, Twitter, Facebook and Instagram)

Project requirements

The customer’s requests were for our team to use their DevOps expertise to: 

• optimize the existing infrastructure and workflows; 

• automate testing and deployment of the new platform features;

• integrate multiple external tools and services;

• migrate the legacy infrastructure to AWS and ensure its ultimate cost-efficiency.

Challenges

The customer required a full-scale transformation of their infrastructure resulted in the scope of the following challenges:

• application containerization and deployment;

• load balancing and autoscaling;

• ensuring security;

• k8s cluster maintenance and troubleshooting; 

• routing and network configuration;

• smart alerting and monitoring implementation.

Results

1. PHP was actively used in developing the main platform at the server and backend application level. We utilized MS SQL Server, Redshift, ElasticSearch and the Lambda architecture at the database storage organization level, focusing on Node.js. 

2. In the working databases, all large accounts and posts for all key social networks are collected and aggregated, with a capacity from hundreds of gigabytes to dozens of terabytes. Search is built on ElasticSearch 6 and consists of 7-8 indexes which contain more than 1 TB of data each and still growing.

3. We started integration with IBM Watson AI services and API. The daily organizing and automating Agile SDLC processes running with GitLab, Jenkins.

4. We prepared the infrastructure for the implementation and coverage of new feature functionality with automated tests;

5. The main application hosting is located in Amazon Cloud using its services (S3, Route53, CloudFront and others).

6. The back-end is a Laravel API. The database is MS SQL, where stored procedures, functions, etc. are actively used.

7. We work with Git to consistently implement Continuous Integration + Continuous Deployment. We use AWS Lambda while working with Node.js.

Final thoughts

With our great DevOps expertise and AWS architecture design best practices, we were able to deliver the expected results and exceed customer expectations. CreatorIQ was fully satisfied with our performance and complimented us in a 5-star IT Svit customer review on Clutch.

Swedish online recipe-sharing platform that also provides grocery and delivery services turned to us with the request to help optimize the back-end performance, enable monitoring and develop new features for their platform.

Case info

Location: Stockholm, Sweden

Industry: Meal planning, online grocery and delivery

Partnership period: May 2019 – ongoing

Team size: 1 Team Lead, 2 DevOps engineers

Team location: Kharkiv, Ukraine

Services: AWS infrastructure performance optimization, monitoring implementation, CI/CD

Expertise delivered: AWS Cloud infrastructure management and optimization, monitoring and alerting implementation, CI/CD configuration

Technology stack: AWS, Kafka, GitHub, Docker, RDS Aurora (MySQL), Elastic Beanstalk, AWS Lambda, EC2, ECS, ELB, CodePipeline, CodeBuild, CodeCommit, AWS auto-scaling groups, CloudWatch, SNS, S3, Github, Varnish, DataDog.

Project requirements

Once the IT Svit team analyzed customer’s needs we specified the following tasks:

• to assess the existing AWS infrastructure;

• to design and implement the failover plan;

• to restructure the current DB setup;

• to finish the CI/CD configuration;

• to optimize the system performance;

• to increase the cost-efficiency of operations;

• to provide detailed developer documentation;

• to provide ongoing system support and updates.

Challenges and Our solutions

The infrastructure the customer operates is split into two regions, in Germany and the US. Both have the test and production environments and run several apps that enable platform functionality. As the product is very popular in Sweden,  so the customer needed us to:

• configure a failsafe CI/CD pipeline for developing and releasing platform updates using AWS resources and services;

• optimize the system performance and reduce the number of AWS services and resources used;

• move the workloads from Germany to the US to increase the response time.

Our solutions:

• we assessed the existing infrastructure and configured AWS VPC peering to optimize resource allocation;

• we moved the production environment to the US and created an Aurora RDS slave replica in the US region;

• we configured AWS CI/CD tools to update both the US and Germany regions simultaneously and replicated the monitoring system from Europe to the US region;

• we also updated the Varnish deployment to change the endpoint based on the region;

• we’ve prepared detailed developer documentation and currently provide project support.

Results

Properly configured AWS services allowed the customer to reduce the expenses on infrastructure maintenance, shorten time-to-market for the platform updates and new features deployment and increase the overall cost-efficiency of the operations.

Final thoughts

The IT Svit DevOps team helped the leading Swedish meal planning platform become more cost-efficient and deliver an uninterrupted positive end-user experience to their customers.

A successful cryptocurrency provider Yallabit contacted IT Svit with the request to develop a fiat-cryptocurrency platform from scratch, manage the infrastructure and provide long-term support of the product.

Case info

Location: Kuwait

Industry: Cryptocurrency exchange

Partnership period: March 2018 – ongoing

Team size: Team of 8 developers

Team location: Kharkiv, Ukraine

Services: Product development, infrastructure management and long-term support

Expertise delivered: Full-stack web development, cloud infrastructure management, CI/CD implementation, monitoring and alerting

Technology stack: JavaScript, Vue.js 2/Vuex/Vue Router, SASS(SCSS), HTML/CSS, Bootstrap 4, Webpack, PHP, Laravel.

Project requirements

Once IT Svit interviewed our client and analyzed Yallabit’s main challenges, we came up with the following tasks:

1. Create an initial infrastructure for running the platform;

2. Develop the end-to-end cryptocurrency exchange solution from scratch;

3. Enable CI/CD configuration;

4. Provide in-depth performance monitoring and long-term support. 

Challenges and Our solutions

The customer’s initial request to set up a cloud infrastructure for their platform grew into a scope of more complex tasks, thus the IT Svit team, when analyzed Yallabit challenges, decided to take these platform development steps: 

• platform design planning;

• UI/UX design and markup development;

• full-stack blockchain software development;

• CI/CD pipelines implementation;

• automated QA and testing;

• API integration for interacting with Kuwaiti bank card gateway KNET;

• monitoring system implementation.

Results

It took us only 4 months to build a profitable software solution from scratch and fulfill all the listed above tasks. The number of users of the Yallabit exchange platform is currently actively growing and we continue to enrich the platform functionality.

Final thoughts

As a reliable Managed Services Provider capable of developing high-quality software within tight deadlines, IT Svit has built a top-notch currency exchange platform and   keeps currently supporting it.

Mobile software developer SBN (Simple But Needed) that offers a set of tools for workplace inspections in government, healthcare and financial industries, needed a senior Python developer and a DevOps specialist to augment their internal team and facilitate the development of various features for their product suite.

Case info

Location: Los Angeles, CA

Industry: Computer software development

Partnership period: January 2019 – ongoing

Team size: 2 Python developers, 1 DevOps engineer

Team location: Kharkiv, Ukraine

Services: Dedicated software development, AWS infrastructure configuration and optimization, monitoring and alerting

Expertise delivered: Dedicated software development, cloud infrastructure optimization, CI/CD implementation, relational DB design and scaling

Technology stack: AWS tools, Python/Django, Django Rest Framework, REST, Docker, Celery, AWS, GKE, Heroku, Prometheus & Grafana, AlertManager.

Project requirements

Once our DevOps leaders discussed issues with the customer and analyzed the situation, they distinguished the following tasks: 

1. to develop functionality for a variety of features;

2. to implement CI/CD processes for software delivery automation;

3. to assist internal specialists in software architecture optimization.

Challenges and Our solutions

Once IT Svit developers and a DevOps engineer had been put on the active SBN team, they did the next:

• studied the existing infrastructure and identified the main bottlenecks SBN team daily faced;

• delivered clean, bugless and flawless codes under deadlines and shared our expertise with the client’s team;

• optimized software development processes utilizing the most effective DevOps practices and consulted the SBN team on practicing DevOps methodology.

Results

IT Svit DevOps engineers:

• reorganized the product structure;

• developed new software features;

• optimized the software delivery process;

• shared expertise and experience with the client’s team.

Final thoughts 

As a trustworthy IT outsourcing company, IT Svit provided reliable services from skilled software engineers and that met and exceeded customer’s expectations. The customer was very satisfied with the professional attitude of our specialists, which resulted in an ongoing partnership

French cryptocurrency company that has developed a bitcoin and cryptocurrency exchange platform and provides blockchain explorer services turned to the IT Svit team with the request to build a system to ensure the wallets cannot contact the blockchain ledger.

Case info

Location: Toulouse, France

Industry: Bitcoin exchange and explorer services

Partnership period: March 2018 – ongoing

Team size: 1 DevOps Team Lead, 1 DevOps engineer

Team location: Kharkiv, Ukraine

Services: AWS infrastructure configuration and optimization

Expertise delivered: Blockchain development, cloud infrastructure management, CI/CD implementation, monitoring and alerting

Technology stack: AWS tools (Lambda, Cognito, RDS, API Gateway), DEX, GitHub, Kubernetes, Terraform, Prometheus, Grafana, ELK stack, OAuth2, Docker, Helm, Vault, Sentry, Peatio, WebSockets.

Project requirements

Once our DevOps team thoroughly analyzed the problem we set our tasks:

1. automate Kubernetes cluster deployment to bare metal servers;

2. automate Kubernetes cluster deployment to AWS;

3. build resource-segregated AWS stack;

4. implement and optimize CI/CD pipelines for product updates;

5. implement monitoring and alerting.

Client’s Challenges and Our solutions

Zobitex platform has a lot of network segregation between different parts of the app and to ensure that wallets are not accessing it directly—they have to pass through a proxy before authentication.

Results

IT Svit dealt with the client’s challenges and here are the results of our work:

• we have prepared Helm charts, Terraform and Kubernetes manifests and other scripts to enable seamless and flawless deployment of Kubernetes clusters to AWS infrastructure; 

• we optimized the system workflows and improved its performance while ensuring the security of operations. We implemented CI/CD pipelines to automate and streamline the delivery of new features and platform operations in production using AWS Lambda, Cognito, API Gateway, RDS and DEX;

• we implemented a monitoring and smart alerting system.

Final thoughts

With Kubernetes cluster seamless deployment to AWS infrastructure, CI/CD pipelines implementation and monitoring system enforcement, the IT Svit team helped Zebitex launch a profitable product and further smoothly update it. These tasks are accomplished, but productive cooperation with Zebitex still goes on.

How IT Svit DevOps Team Containerized a Web Conferencing App

A global company, the creator of the conferencing tool used by 19,000 customers in 100 countries, contacted the IT Svit team with the request to set up a Kubernetes cluster and containerize their application. Also, they wanted us to enable CI/CD pipelines for updates and monitoring. 

Case info 

Location: Barcelona, Spain; Hamburg, Germany; Harrisburg, Pennsylvania, USA.

Industry: Web conferencing software

Partnership period: July 2019 — January 2020

Team size: 1 team lead, 2 DevOps engineers

Team location: Kharkiv, Ukraine

Services: Cloud infrastructure design and optimization, database performance optimization, monitoring implementation, CI/CD implementation, app containerization, API integration

Expertise delivered: App containerization and container management, cloud infrastructure management, monitoring implementation, CI/CD configuration,

Technology stack: AWS, GitHub, Jenkins, Ansible, PHP, MySQL, MariaDB, Apache, Nginx, Terraform, Docker, Prometheus, Grafana, AlertManager

Project requirements

Our customer needed help with: 

• improving their infrastructure by splitting the platform into a set of microservices with containerized apps;

• building and configuring the Kubernetes cluster for the platform;

• enabling monitoring of the platform. 

Our solutions 

The work process included: 

• challenges around persistent data such as DB, file uploads and user-generated content resolving:

• Redis for PHP sessions configuring;

• Dockerfiles to structure the platform’s Docker Registry creating;

• config data to PHP container runtime passing;

• Kubernetes setting up;

• monitoring and alerting pipeline enabling;

• system dependencies managing.

Results

Here are our results:

1. By deploying the platform consisting of microservices with containerized Docker apps to a Kubernetes cluster, the IT Svit DevOps team enabled security, load balancing and auto-scaling of resources;

2. By implementing CI/CD processes, we allowed our customer to update various platform components independently and their performance is not affected by other operations;

3. In-depth monitoring ensures the customer’s team keeps a hand on the pulse of the project at all times and the TTR has shortened significantly.

Final thoughts

Our talented DevOps team used the most effective and innovative DevOps practices and tools to ensure error-free deployment, updates and maintenance of enterprise-grade conferencing software used by thousands of people daily.

How IT-Svit Improved Insights Processing Pipeline at Data Analytics Platform

Co-founder & COO at a real-time Big Data analytics platform got in touch with our team. The platform provides an AI-powered methodology to get the most out of your business data that helps dig valuable insights out of the data piles your business has access to.

Case info 

Client’s Location: Los Angeles, CA

Client’s Industry: Business Intelligence and Big Data analytics

Partnership period: December 2018 — current time

Team: 1 Team Lead, 2 DevOps engineers, Big Data engineer

Team location: Kharkiv, Ukraine

Services: Cloud infrastructure design and optimization, database performance optimization, Machine Learning model training, monitoring implementation, CI/CD implementation

Expertise delivered: App containerization and container management, ML training and management, cloud infrastructure management, monitoring implementation, CI/CD

Technology stack: AWS, Coordinator, Alertmanager, G Suite, GitHub, Jenkins, Kibana, OpenVPN, Prometheus, Spinnaker, Superset.

Project requirements

The customer’s Big Data system run on AWS infrastructure needed IT Svit to: 

1. upgrade Druid instance;

2. implement enhanced system monitoring and alerting with Amazon Cloudwatch, Prometheus & Grafana and AlertManager;

3. configure a staging environment with Kubernetes, Ansible, Docker, Apache Kafka, Hadoop and Druid, Postgres, Node, AlertManager and Prometheus & Grafana;

4. plan and build a CI/CD pipeline using Jenkins to improve the insights processing pipeline.

Challenges and Our solutions

Our client constantly experienced such uncomfortable challenges that significantly slowed down product growth:

• Marathon infrastructure was overwhelmed with ever-growing requests;

• Big Data input sources were slow and the obsolete Druid version needed updating;

• The complex backend structure was not able to provide info about the system’s performance.

Our solutions: 

• to set up Airflow for Data Analysis automation;

• to update Druid and Spinnaker to the latest versions;

• to set up the ECS cluster and move apps from Marathon to Amazon EKS;

• to install and configure Jenkins pipelines;

• to set up monitoring and notifications with the help of Prometheus, Grafana and AlertManager.

Results

IT Svit team accomplished all the tasks in such an order: 

1. We optimized the architecture of the system by moving all the apps from Marathon to Amazon EKS. This allowed us to reduce the Druid cluster upgrade time to 15 min;

2. We found and fixed the issue that caused application workers to restart frequently on legacy clusters;

3. We configured the CI/CD pipeline using Jenkins to automate daily cluster operations.

4. We connected Spinnaker to Prometheus with the help of AlertManager. System abnormal behavior is immediately alerted to Slack and is being resolved much quicker now.

Final thoughts

It Svit solutions like Druid and Spinnaker updating, moving from Marathon to EKS, Jenkins pipelines configuration and monitoring system installation allowed our client to operate their Big Data analytics platform more effectively and much faster.