Lead Engineer at Rezolve AI · Kolkata, India
Open to freelance projectsI build voice AI and agent systems that hold up in production.
Lead Engineer at Rezolve AI. I design real-time voice pipelines, multi-agent architectures and the cloud-native backends behind them, with a bias for low latency and sensible running costs.
Thursday at 4:30 with Dr. Rao? Booked. Want a reminder?
// a voice turn, end to end
What happens after you say hello.
- 01
You speak
Audio streams in over WebRTC
- 02
LiveKit
Real-time media transport
- 03
Speech-to-text
Azure Speech, streaming
- 04
LangGraph agent
Tools, RAG and business systems
- 05
Text-to-speech
Azure neural voice, streamed back
// what I work on
Voice, agents, and everything underneath.
Real-time Voice AI
Streaming speech pipelines on Pipecat, LiveKit, WebRTC and Azure Speech, tuned for latency, recognition accuracy and natural interruptions.
Backend architecture
Django REST Framework and FastAPI services on PostgreSQL, MongoDB and Redis, with async processing that stays maintainable as it grows.
Agentic AI & RAG
Multi-agent systems with LangGraph, tool calling and retrieval, connected to business systems through MCP and reusable SDKs.
Cloud & DevOps
Azure, AKS, Docker, Kubernetes, Helm and CI/CD pipelines that make deploys repeatable and boring, in the good way.
MLOps
Packaging, versioning, deployment and monitoring workflows that take models from a data scientist's notebook to production.
Mobile & full-stack
Flutter apps with GetX on top of Django and FastAPI backends, from reusable UI components to CI with GitHub Actions.
// selected work
A few things I've built.
Have a voice or agent problem worth solving?
I take on a small number of independent projects, and I’m always happy to talk shop.