Python Developer in the R&D department. I specialize in Python development, prompt engineering, and building AI-powered tools using LLMs. Key projects:
– Developed a Telegram chatbot for internal knowledge search. Implemented a custom RAG pipeline (embedding-based retrieval + prompt tuning), later migrated to OpenAI Assistant for improved accuracy.
– Created a Telegram chatbot for end-users to answer 1C-related questions, reducing the load on tech support by handling frequent queries.
– Selected an LLM (DeepSeek R1:14B) to detect critical emails for the QA team. Evaluated models with Ollama and prompt tuning, optimized for dedicated RTX 3070 Ti resources.
– Deployed a Flask-based grading service for Python assignments submitted by students on the company’s learning platform, using OpenAI API.
– Built a Flask server for students career guidance: the company’s platform sends API requests on the company’s learning platform. Requests are processed and evaluated via LLM.
