Denys – Python, Django, Flask

This platform was a fully featured video streaming service offering live broadcasts, recordings, and VOD content, targeting both business (B2B) and consumer (B2C) markets. For B2B partners, it provided a white-label solution—enabling them to quickly launch their own branded streaming service—while the B2C version delivered content directly to paying subscribers. A standout feature was its integrations with leading media libraries, including Disney, Apple Tv, Amazon Prime, and national broadcasters, significantly expanding the available content catalog. Additional key features included a robust streaming architecture for high-traffic demands, secure payment and subscription tools, and a user-friendly interface to ensure seamless customization and viewer enjoyment.

• Integrated major media providers (including a national public service broadcaster) into a Ruby on Rails CMS, securing five new partnerships and boosting company profits.
• Enhanced the white-label product API by pulling content ratings from IMDb with Python and CherryPy, reducing user selection time by 25% and unlocking new partnership opportunities.
• Improved product performance by 15% through establishing a unified code style (Rubocop, Ruff, pylint, mypy) and automating checks with GitHub Actions and Jenkins.
• Accelerated development cycles by implementing a CI/CD pipeline in Jenkins and GitHub Actions, integrating code quality checks and tests, and cutting pull request turnaround by 50%.

This platform was a marketplace for used items—from cars and real estate to smaller gadgets like laptops and TVs. Unlike traditional buy-and-sell platforms, it introduced a custom credit scoring system that replaced the need for third-party credit agencies. Buyers could secure financing directly through the platform, with sellers acting as lenders under a personalized loan contract. This unique approach simplified the payment process, making it easier for users to buy and sell high-value items.

• Devised an automated contract generation system using Google Docs API and AWS Lambda, reducing deal closure time by 67% and processing over 150 contracts monthly.
• Engineered integrations with key financial APIs (TransUnion, Alloy, Plaid, Treasury Prime) using Python, streamlining KYC checks, credit scoring, and payment flows.
• Boosted user engagement and retention by 35% through Firebase integration for real-time notifications and OAuth authentication.
• Implemented Agile development practices (Scrum ceremonies, sprints, stand-ups), cutting product delivery time by 25% and raising team efficiency by 30%.
• Automated software deployments using GitHub Actions for CI/CD, lowering deployment errors by 34% and minimizing downtime.

This platform was built for hospitals to enhance social interactions and streamline management for long-term patients. Designed with patient engagement in mind, it integrated video calls via Zoom, offered custom image and video processing for secure messaging with relatives, and supported efficient communication between patients and hospital staff. By combining social features with administrative tools, the platform aimed to improve patient well-being, reduce isolation, and simplify hospital workflows.

• Directed the technical design for a hospital communication platform using Go, Python, Flask, and RESTful microservices, increasing development efficiency by 15% and ensuring a successful product launch.
• Managed client communications and proactively addressed shifting requirements, elevating customer satisfaction by 20% and maintaining timely deliveries under Agile processes.
• Designed and delivered complex features (push notifications, chat, media processing) in Python, Django, and FastAPI, boosting user engagement by 45% and enhancing overall functionality and security.
• Conducted regular code reviews to mentor colleagues, improving coding standards compliance and reducing defects by 20%.

This service was a ride-hailing platform specifically tailored for cargo transportation. Initially designed for B2C customers, it later pivoted to a B2B model, offering businesses a streamlined way to schedule and manage freight deliveries. Its serverless architecture relied on AWS Lambda functions and API Gateway, with a sophisticated, AI-driven route planner at its core. By optimizing delivery paths in real time and operating in a stateless manner, the platform efficiently handled dynamic workloads and ensured a seamless cargo transport experience for both companies and end-users.

• Optimized PostgreSQL queries via in-depth profiling, reducing execution time from over two minutes to 20 seconds and enhancing overall scalability and user experience.
• Implemented a complex user scope management system with AWS Lambda and API Gateway, strengthening product security and enabling a successful new product launch.
• Led the technical design and development of key features, collaborating with the Product Owner to propose innovative solutions, secure approvals, and improve team efficiency.

This platform was designed to completely eliminate paperwork in the car-buying process. It allowed buyers to apply for vehicle loans in a fully digital format, while automatically calculating taxes and fees for any payment method—no matter how complex. Built for individuals and dealerships looking to streamline auto purchases, the platform’s main features included a paperless loan application workflow, real-time fee calculations, and a user-friendly interface that expedited the entire transaction process.

• Engineered a high-load, automated data pipeline using Go and AWS (Lambda, S3, SNS), boosting processing efficiency by 40% and handling a 55% spike in data volume with zero downtime.
• Migrated SOAP-based APIs from IBM API Connect to RESTful APIs on AWS API Gateway, redesigned them for REST architecture, and reduced response times by 35%.
• Led a team of peers to overcome technical challenges, improved collaboration, and increased overall productivity by 15%.
• Planned and executed the migration of three critical RESTful services to Python 3.8 for over 100k users, enhancing performance by 12% and strengthening security.
• Upskilled in React to resolve vital integration issues in a back-end-centric environment, applied fixes across front-end and back-end components, and improved system functionality and user experience.

This platform was designed for businesses looking to collect and analyze feedback on their client services. It featured a multi-level user management system—complete with group-based permissions and staff oversight—a dynamic questionnaire builder for creating custom surveys, and an automated data aggregation engine. By presenting results through comprehensive analytics dashboards, the platform empowered businesses to gain real-time insights and make data-driven decisions to enhance customer satisfaction.

• Migrated the entire codebase from Python 2 to Python 3, improving performance and long-term maintainability.
• Recreated the migration structure from scratch, resulting in cleaner database schemas and smoother upgrade paths.
• Containerized all services using Docker, streamlining deployments and ensuring consistent development environments.
• Overhauled legacy logic in line with the latest Django guidelines, bolstering reliability and future-proofing the platform.
• Updated dependencies to their most recent versions, reducing security vulnerabilities and leveraging new features.
• Introduced automated creation and uploads of database backups to Amazon S3, securing critical data and enabling swift disaster recovery.
• Integrated Slack for real-time notifications, enhancing collaboration and speeding up issue resolution.
• Developed a custom deployment script, cutting manual effort and lowering deployment errors.
• Fixed multiple client-reported bugs, significantly improving overall stability and user satisfaction.

This platform was created to streamline the storage and analysis of scientific research data. Intended for in-house use by scientists conducting various studies, it handled the upload, processing, and secure storage of experimental results. The system also featured automated analytics for multiple research types and a robust user permissions framework, ensuring that data remained accessible only to authorized personnel. Finally, it exported processed insights into human-readable file formats, simplifying collaboration and enabling faster decision-making within research teams.

• Designed the core architecture for scientific data processing.
• Implemented complex database models to store and manage research results efficiently.
• Developed the back-end logic for a user dashboard, providing secure access to real-time data.
• Created an import/export module for research findings, streamlining data sharing among teams.
• Fixed critical bugs, improving platform stability and user satisfaction.
• Increased unit test coverage to enhance overall reliability and reduce regressions.

This plugin for the open-source graphic editor Inkscape applied specialized mathematical transformations to images, primarily for computer vision research. It was developed to help scientists train their models and generate visual examples for scientific papers. By integrating seamlessly with Inkscape, the plugin enabled real-time previews of transformed images, streamlining the workflow for researchers who needed high-precision, reproducible image modifications in a user-friendly environment.

• Architected the plugin framework within Inkscape, enabling advanced image transformations for scientific research.
• Implemented and optimized the mathematical transformation logic, ensuring accurate and reproducible results for model training.
• Covered the codebase with comprehensive unit tests, increasing reliability and simplifying future maintenance.