- 🤖 AI & LLMs: OpenAI (GPT-4, GPT-5), Claude, LangChain, LangGraph, LLMs, Local LLMs, RAG, Prompt Engineering, AI Agents, Agentic AI Development
- 🎙️ Voice AI: Vapi.ai, TTS, Whisper, AI Voice Agents
- ⚡ Automation: n8n, Make.com, Zapier, AI Workflow Automation, custom webhooks
- 👨💻 Backend: Node.js, NestJS, FastAPI, TypeScript, Python, REST APIs, SSE
- 🔭 Frontend: Next.js, React, Tailwind CSS
- 🗄️ Database: Supabase (PostgreSQL), MongoDB, SQL Server, Chroma, Redis
- ☁️ Infra: Vercel, Docker, Google Cloud, Azure API Management, Azure Key Vault, CI/CD
- 🔧 CRM & Tools: GoHighLevel (GHL), HubSpot
A full-stack RAG (Retrieval-Augmented Generation) learning project built with the modern Python AI stack. Upload documents, ask questions, and watch the entire pipeline execute live - with technique toggles to compare retrieval strategies.
| Layer | Stack |
|---|---|
| Frontend | Next.js 16 + Tailwind + shadcn/ui - Vercel |
| Backend | FastAPI + LangChain + LangGraph - Fly.io / Railway |
| Storage | PostgreSQL + Chroma (vectors) + Redis |
Pipeline: query -> [classify] -> [retrieve: naive | hybrid | multi-query | HyDE] -> [rerank] -> [generate + citations] -> SSE stream -> UI





