I design and build production-ready AI/ML systems with a focus on LLM-powered applications, RAG pipelines, multi-agent AI workflows, and speech-to-speech systems. I love turning cutting-edge AI research into real products that scale and deliver impact.
- Data Science Engineer @ 10Pearls | AI/ML, LLMs, RAG & Speech-to-Speech Systems
- AI/ML Solution Executive @ Prudential Solutions | Building multi-city, multi-agent chatbots and real-time AI workflows
- Passionate about deploying AI in production, designing microservices, and solving real-world problems with ML
- Supports 220+ languages with high-accuracy speech recognition and synthesis
- Low-latency streaming via OpenAI SDK and FastAPI backend
- Frontend built with React/TypeScript for seamless real-time interactions
- Frontend Repo | Backend Repo
- Predicts Air Quality Index up to 3 days in advance using ML & Deep Learning
- CI/CD-enabled MLOps pipeline with hourly feature updates & daily model retraining
- Interactive Streamlit dashboard for real-time monitoring
- Live App | Repo
- Production-grade AI agent platform with multi-provider LLM routing (OpenAI GPT-4o + Anthropic Claude)
- Voice I/O pipeline with real-time STT via Whisper and streaming TTS via ElevenLabs
- RAG knowledge base (ChromaDB) with WebSocket token streaming; deployed live at aigilityx.com
- Live App | Repo
- Advanced System Design for AI products
- Large-scale multi-agent AI orchestration
- Optimization of real-time speech pipelines
