DATA212

Challenges: ​The main goal was to create an ML model to assess risk and identify potential and current tenants with the lowest and highest delinquency risk for residential rentals.

Results:

• Our team has created a machine learning algorithm that is specifically designed to evaluate the bank transaction data of potential renters. The algorithm can analyze spending patterns and identify key indicators that accurately reflect the applicants' true level of income and their ability to make rental payments.
• The solution assists our client in gaining a comprehensive understanding of each applicant's financial situation. Our algorithm constantly learns and improves to provide the most precise and dependable results.

Technologies:

• Core stack - AWS Lambda, S3 Bucket
• Database - PostgreSQL

Challenges: The main goal was to develop a risk monitoring and assessment system associated with mortgages taken out by individuals, as they can pose a threat to both the borrower and the overall financial system.

Results:

• We have developed an algorithm that analyzes bank transactions of mortgage holders to detect potential risks. 
• Any identified risks are reported to the regulator, along with an expected timeline for non-payment based on a loan equity evaluation.

Technologies:

• Core stack - AWS Lambda, S3 Bucket
• Database - PostgreSQL

Challenges: The main challenge was to develop data analysis and ML strategies for clients and partners of TechUnit from different industries and countries. 

Results:

• We prepared customized, data-driven solutions that propel operational efficiency, enhance customer experience, open new market avenues, and make the business scalable. 
• Throughout our collaboration, we have consistently embraced new challenges, paving the way for strategic expansion within our clients' enterprises. 
• Our team has been at the forefront of crafting custom machine learning models that enhance our clients' product offerings and business verticals, thereby catalyzing growth.  
• The satisfaction of TechUnit's clients with the outcomes achieved is testament to our success. This satisfaction has translated into a continuous stream of new opportunities for us, reinforcing our commitment to driving innovation and delivering value.
  
  Technologies:
• Core stack - Deployment Services: EC2 Dashboards: Plotly
• Database - PostgreSQL

Challenges: The primary objective was to collect data from a CRM system that has no API and turn it into comprehensive and informative dashboards.

Results:

• Our engineering set up a mechanism to extract data from the client’s CRM system. Following this, we formulated a comprehensive analysis strategy to discern patterns and distill crucial insights from the amassed data.
• A tailor-made dashboard was then developed, offering intuitive and accessible visual representations of the data. This dashboard is designed to empower the client to make informed, data-driven decisions based.
• In the culmination of our efforts, we ensured the dashboard's reliability and scalability through the judicious selection of tools and technologies. This ensures its continued efficacy and adaptability to the client's evolving business needs, thereby supporting sustained growth and transformation.

Technologies:

• Core stack - AWS Lambda, S3 Bucket
  Database - PostgreSQL

Challenges: The main goal was to personalize the learning experience by suggesting trial lessons that match the unique needs and interests of each individual, while the challenge was to offer an automated solution that could allow the client to achieve the set goal.

Results:

• Our team has developed a machine learning model that predicts the probability of a student purchasing a trial lesson based on their past behavior and demographic data. This model can be used to automatically suggest trial lessons to students who are likely to buy them.
• Model is also used to improve customer acquisition by identifying the potential customers, who are more likely to convert into high-LTV users.
• To train the ML model, we used data from the online school's learning management system (LMS), student profiles and CRM. We employed ML algorithms to generate scores for each trial lesson indicating the likelihood of a student purchasing it.
• As a result, the scoring model provided personalized recommendations for trial lessons, which increased the chances of student engagement and conversion to paid subscriptions. This, in turn, enhances revenue and student satisfaction for the online school.

Technologies:

• Core stack - Sklearn, Python, Pandas, S3 bucket, AWS Lambda
• Database - PostgreSQL

Challenges: The main goal was to gather data from a solar power plant, store it in a database, and design a Tableau visualization to fit the client's specifications.

Results:

• Leveraging our proficiency in web scraping and data engineering, we efficiently extracted and archived information from the solar generation facility.
• A bespoke database was meticulously crafted to ensure the secure retention of data, alongside guarantees of its reliability and scalability. 
• Our specialists then devised an interactive dashboard within Tableau, showcasing essential performance metrics such as energy output, operational efficiency, and downtime instances. 
• Furthermore, we constructed an intuitive user interface, facilitating real-time monitoring of the solar station's performance by the client. 
• Ultimately, this initiative empowered the client to formulate decisions rooted in data, driven by the comprehensive insights delivered through the dashboard.
• At the same time, the transparency into the demant/supply dynamics allowed the business to build processes that made it more resilient to energy production and consumption fluctuations through strategically optimizing operational efficiency.

Technologies:

• Core stack -  AWS Lambda, S3 Bucket
• Database - PostgreSQL

Challenges: The main goal was to delve into customer behavior and usage trends, aiming to make well-informed decisions focused on the consumer within an expanding market.

Results:

• Initially, we carried out a detailed examination of the client's business, segmenting it into specific areas of interest. This was to fully grasp the necessity for certain analyses and their potential impact on fulfilling business objectives. Our comprehensive analysis spanned various aspects, including crucial financial insights, comparisons of monthly and annual trends, evaluation of marketing channels, segmentation, sales performance, personnel analytics, and other pertinent reports.
• After gathering the necessary data from the internal database and CRM systems, we proceeded to develop accessible reports via Tableau. These reports featured significant metrics, informative visualizations, and most critically, actionable data. This approach enabled stakeholders to make well-informed decisions for the business.
  
  Technologies:

• Core stack -  AWS Lambda, Amazon RDS (postgreSQL), S3 Bucket (dev-support storage)
• Database - PostgreSQL