
AI-Powered Support Desk Analytics Platform
A full-stack AI application that helps support teams analyze customer-service tickets without writing SQL. The system ingests a CSV dataset, validates the schema, stores the records in SQLite, answers natural-language questions, and highlights potentially problematic tickets.
The application uses an LLM only to understand the user's question and produce a restricted analytics plan. Pydantic validates that plan, while deterministic Python code performs the actual calculations. This improves reliability and prevents the model from directly generating unsafe or unpredictable SQL.
Tools: Python, FastAPI, Streamlit, SQLite, Pydantic, Ollama (Qwen2.5:7B), Groq, Uvicorn, Docker





