Portfolios
AI-Driven Knowledge Base & RAG Bot
AI-Driven Knowledge Base & RAG Bot
I solved the challenge of unreliable AI responses by building a "closed-loop" RAG system. Using PGVector and OpenAI embeddings, I created a Telegram bot that strictly answers based on a verified internal database, eliminating hallucinations. The entire ecosystem—from web scraping to database updates—is fully automated via n8n, ensuring a self-sustaining and high-precision search portal.
Tech Stack: Django, n8n, PGVector, OpenAI, and Python.
Smart News Aggregator & Community CMS
Smart News Aggregator & Community CMS
This project streamlines information flow by combining AI automation with community reporting. Built using a Django-Bootstrap stack, the system handles end-to-end data processing: from 24/7 automated scraping and AI-driven summarization to a dedicated local CMS. It ensures the community stays informed with curated, easy-to-read content delivered in real-time.