Hansraj Singh Thakur · AI Engineer · Seattle
I build AI systems that ship.
Fifteen years from embedded hardware to autonomous agent organisations. Now building agents, voice interfaces, and full-stack products. In public, every day.
Showreel
HANSRAJ.OS, in 24 seconds.
Boot terminal, a 25-agent constellation, live GitHub stats. Built with Remotion. Unmute for the soundtrack.
About
A builder at the edge of what's possible.
I build intelligent systems end to end. From embedded hardware in the earliest years to autonomous agent organisations today.
Craft, velocity, and the honest feeling of pushing something live at 2AM and watching the uptime graph hold.
Origin story
Started with embedded systems: Kinect, computer vision, surveillance hardware, where the only feedback loop was whether the thing blinked. Grew into systems architecture, then platform scale, then a multi-year GenAI transformation replacing brittle legal workflows with RAG across 108 data sources.
Now
Builder mode. Running a 25-agent autonomous AI organisation that commits to GitHub daily. Shipping macOS tools for local LLMs. Prototyping the agentic web on nights and weekends. Everything in public.
Experience
Fifteen years, one compass.
Engineer to technical program lead to independent builder. Roles and outcomes, 2011 to now.
- Led 20-person dev team delivering enterprise LOB applications end-to-end
- Designed and shipped complex data lake solutions from scratch
- Hands-on with embedded systems: Kinect, computer vision, surveillance hardware
- Designed cross-geo enterprise data architectures spanning multiple continents
- Built requirement-gathering frameworks that cut delivery cycle time by 30%
- Grew from individual contributor to leading cross-functional global teams
- Architected a 45TB analytics platform delivering real-time insights to 1.1M users across 130 countries
- Led 8 Technical PMs + 80 engineers, $10M annual revenue accountability
- Optimized big-data query performance and drove hyper-scalability design
- Spearheaded multi-year GenAI initiative: RAG-based LLMs across 108 data sources
- Saved $50M annually; boosted legal team productivity by 40%
- Led 4 Sr. TPMs + 150 engineers, $20M annual revenue, 99.9% uptime SLAs
- Running a 25-agent autonomous AI org (Indian Avengers) with daily GitHub commits across all repos
- Built AgentDate — agent discovery/registration platform; AgentMesh — real-time multi-agent visibility layer
- PMChat, Mission Control dashboard, OllamaBar — shipping in public with Claude Code + OpenClaw
Work
Selected work, shipped in public.
Five of many. Every one of these is a real repo you can read.
AgentMesh↗
Real-time visibility and coordination layer for multi-agent AI systems. See them coordinate; debug them at speed.
PMChat↗
The AI first-mate for product managers. Turns scattered requirements into structured specs.
OllamaBar↗
Local LLMs tucked into the macOS menu bar. Zero-friction inference from wherever you are.
Mission Control↗
A 25-agent war room. Real-time briefings, task dispatch, and fleet telemetry for the autonomous dev org that ships to GitHub daily.
InterviewAgent↗
An intelligent automation system that lets candidates apply to hundreds of positions simultaneously — resume tailoring, cover-letter generation, and submission handled end-to-end by agents.
Skills
End-to-end, not end-of-stack.
AI & Machine Learning
Frontend
Backend
Mobile & Desktop
Hardware & Embedded
Now
Live at the commit level.
What's moving this month, on a daily cadence.
AgentMesh↗
Real-time visibility and coordination layer for multi-agent AI systems
PMChat↗
AI-powered product management — turns scattered requirements into structured specs
Mission Control↗
War room for the 25-agent org — briefings, task dispatch, and fleet telemetry
Contact
Want to learn more?
Connect with me. I'm on GitHub and LinkedIn, and email always works. Based in Seattle.