# Development of a comprehensive DaaS platform providing real-time, automated data on drugs, diseases, clinical trials, and financial trends. Enhancing data standardization and cross-referencing using semantic technology, aiding investors, portfolio managers, analysts, and quants in making informed decisions- Improving backend infrastructure for cleaner, scalable, and maintainable code- Built robust pipelines to process data from multiple external APIs, ensuring accuracy and timeliness.Environment: Python, Django, Airflow, Celery, RabbitMQ, AWS S3, AWS RDS # Developed a chatbot app powered with LLMs and RAG for an HPC system, streamlining HPC usage for scientists. Spearheaded the creation of a backend infrastructure, optimizing user message history management. This AI-powered initiative significantly aided scientists in managing clusters with Slurm and reduced the workload on helpdesk support. - Implemented data persistence for users’ message history, ensuring seamless continuity and better user experience. - Developed a context length limiter mechanism using a sliding window algorithm, reducing model token usage and significantly cutting company costs. - Conducted in-depth research into various vector databases, enhancing the precision of the RAG mechanism. Environment: Python, FastAPI, PostgreSQL, LangChain, OpenAI API, Slurm, AWS Parallel Cluster# Developing a proprietary LIMS application to effectively streamline sample management; the advantage of this product over existing market solutions lies in the ability to simulate and customize any laboratory environment. - Implementing key functionalities for efficient sample manipulation- Designed and integrated a GPT-based assistant to accelerate user interactions and improve user experience- Refactor critical elements of the back-end infrastructure to ensure cleaner, more maintainable, and scalable code.Environment: Python, FastAPI, GraphQL, PostgreSQL, Azure