Stack
The tools, frameworks and models I use every day to take AI systems from a notebook to production.
LLM & GenAI
Orchestration, agents and the glue between models and real systems.
- LangChain
- LangGraph
- LangSmith
- LiteLLM
- RAG
- Agentic AI
- Multi-Agent Systems
- MCP (Model Context Protocol)
- Tool Calling
- Prompt Engineering
- LLM Evaluation
- Embeddings
Voice AI
Real-time speech pipelines where every millisecond counts.
- STT / TTS Pipelines
- ASR
- VAD
- Speaker Diarization
- Deepgram
- Whisper
- DeepFilterNet
- LiveKit
- WebRTC
- Twilio
- Exotel
- Acefone
- FFmpeg
Models & providers I’ve shipped or benchmarked
- OpenAI
- Groq
- Gemini
- Anthropic
- Deepgram
- Sarvam
- Murf
Backend
- Python
- FastAPI
- Async Python
- REST APIs
- WebSockets
- Microservices
- Real-Time Systems
- Docker
Data & Retrieval
- PostgreSQL
- Redis
- MongoDB
- MySQL
- SQLite
- FAISS
- Pinecone
- ChromaDB
- BM25
- Microsoft Presidio
Machine Learning
- PyTorch
- Scikit-learn
- Deep Learning
- NLP
- Computer Vision
- Transfer Learning
- Reinforcement Learning
- OpenCV
Also
- SQL
- JavaScript
- TypeScript
- React
- Streamlit
- Git
- Google Cloud
- Jupyter