UplineSoft

Customer and task

Our client approached us to ensure the quality of a video surveillance system developed based on artificial intelligence technologies.

Solution

As part of working on the customer's project, we were able to complete the following tasks:

• Developing and writing autotests for existing and later new product functionality. This reduced the amount of manual testing and eased the transition into automation.
• Test case coverage, testing of new and existing functionality.
• Writing scripts for load testing.
• Testing of project documentation.

Results

We successfully completed the assigned tasks for ensuring the quality of the system within the specified deadlines and also provided several suggestions for the development of automated testing.

The client implements a ticketing technology platform with its own API for carriers and travel companies.

Objective

Audit the platform for vulnerabilities from external threats

Solution

At the startup, we studied the product as there was no actual documentation ready yet.

The methodology of the work was as follows:

• We collected information and immersed ourselves as much as possible in the processes of the platform.
• We determined the direction of attacks and current vulnerabilities.
• We identified and categorized vulnerabilities.
• Based on the information received, we provided recommendations.

Results

We presented the client with a list of recommendations to improve the security of the platform.

Customer

Our client is developing a services marketplace with a mobile application for customers.

Objective

As part of the in-house team, complete the active development phase and prepare the minimum viable product version of the application (MVP) for release.

Solution

The project was in the active development phase, and we were involved in the creation of every feature. For example, we implemented the registration state recovery: even if the user interrupts the registration process and exits the application, later they can return to the same point. We are also responsible for implementing a biometrics module and a customer profile.

Results

Previously, the developers' efforts and time were spent on support and development of the main marketplace platform, so the deadlines were constantly postponed, with only the registration stage of the application implemented. We helped reduce technical debt and added key features for the release and further development and promotion of the product.

Our client, a chain of electronics stores, uses an IT system to analyze business performance and manage workflows.

Objective

Development of new functionality of CRM-system and mobile application

Solution

We completed a series of tasks to reduce code coupling, optimize, and refactor.

We updated the application build tools and implemented CI/CD using Fastlane + Match for continuous integration and publishing of the application.

Results

We completed the assigned tasks for optimizing the performance of the mobile application and integrated build automation for enterprise. New functionality is still being developed at this time.

Our client is a chain of over 30 restaurants with their own web-based food delivery application. Our goal was to ensure that the application could support 10 times the current load.

Objective

Conduct technical audits of CI/CD infrastructure and processes for fault tolerance, security and manageability.

Solution

Working closely with the customer's team, we identified a number of solutions in use that were not in line with common infrastructure practices (anti-patterns). After that, together with the client, we categorized the defects found by criticality and began to develop a plan for their elimination.

We proposed two solutions:

● try to correct deficiencies of the existing infrastructure

● migrate to a more suitably arranged infrastructure

Our client chose the second option, after which we examined their internal processes and proposed a move from self-managed to cloud-managed infrastructure.

In creating the new architecture, we were not looking for a short-term solution to a current problem, but a method that would work for the future. We proposed the Infrastructure as Code (IaC) approach. It will allow the customer to better manage their environments, and if necessary, change cloud providers at a lower cost.

Results

The audit allowed us to identify weak points in the infrastructure setup and develop a plan to improve the system. By working cohesively with the client's team, we were able to help them achieve their business goals.

A complex solution uses mathematical algorithms to visualize a borehole in 2D space. The desktop app analyzes rock structure for geologic horizons map reconstruction and improves accuracy of drilling.

Customer’s benefits:
— Improves drilling performance, adhering to a certain geological layer.

UPlineSoft challenge:
— Stress tests to fasten software performance while transmitting huge amounts of data
— 1000 simultaneous participants
— 3 years of product support

Web blockchain-based system allows users to set prices, bill and pay for electricity.

To access the system, a user gets a device (a smart meter) with preinstalled software, the meter address, and cryptocurrency wallet keys.

Functionality overview:
— Create and deposit the wallet
— Set the price and bill for energy
— Pay for consumed electricity
— Open / Close smart contract
— Calculate the charge
— Operator dashboard: monitor device dynamics, user management

Web platform for hiring a person for instant interpretation.

Clients can create jobs; score, invite and book interpreters, control job time, pay for the job via:
— PayPal
— Credit cards
— Direct payment, bank transfer

The functionality of payment guarantee and money return is provided.

UPlineSoft challenge:
— GDPR requirements
— CRM integration, telephony
— Adaptive layout

In general it is a system of docflow automation in healthcare centers in the UK with a speech recognition module.

The System stores documents, converts audio into text, prepares and prints out letters and reports.
— The system is used by 45,000 health workers daily.
— 4,000,000 medical disposals were transcribed during 2020.

The system predicts and prevents various negative incidents on 600+ servers in the customer's infrastructure by detecting anomalies when analysing logs.

UPlineSoft challenge:
— Patterns of logs behaviour were determined and analytical models were built based on them
— Incidents on the server array 10-15 minutes were predicted before their occurrence and warn the personnel.

A solution for insurance companies that automates the process of working with a patient.

The system makes a decision based on a set of machine learning models:
- identification of a non-insured case, treatment with the same diagnosis;
- recommending procedures from the insurance coverage list;
- how the patient will feel after 30, 60 and 90 days.

Achievements in 6 months:
- integrations with multiple medical systems;
- ML models are developed, tested and continuously trained.