I engineered and developed DocQuery AI, a modern full-stack AI SaaS platform leveraging Retrieval-Augmented Generation (RAG) to enable enterprises to analyze complex documents, contracts, and financial reports with instant, page-verified citations.

Key Roles & Contributions: • End-to-End Development: Designed and built the entire application using React, TypeScript, Tailwind CSS, and Supabase. • Semantic Search Engine: Implemented vector similarity search powered by PostgreSQL PgVector and cosine distance metrics for high-speed chunk retrieval. • RAG & LLM Integration: Integrated Google Gemini API with strict zero-hallucination guardrails, ensuring factual responses strictly grounded in uploaded files. • Modern UI/UX: Crafted an Apple/Linear-inspired dark UI supporting full bilingual (Arabic/English, RTL/LTR) responsiveness. • Enterprise Security & Cloud Deploy: Enforced Row Level Security (RLS) policies for multi-tenant isolation and deployed on Vercel infrastructure.

Live Demo: https://docquery-two.vercel.app/