Nine years running enterprise software from the customer's side — $2M+ ARR, India's biggest brands, not a line of code among them. Then he stopped advising on AI and started building it. Thirty days later, he'd shipped an autonomous harness that plugs into any workflow and gets smarter every session — arriving, from first principles, at the same patterns the research world was just naming: context engineering, layered memory, compound systems.
Nobody builds an autonomous AI system in a month by accident. The leap only looks sudden — here's the road that made it inevitable.
Harness Commander is an intelligence layer that plugs into any workflow — bringing persistent memory, encoded judgment and self-correction — and compounds with every session. Customer Success is simply where it's being proven first.
Proven in the domain he knows cold — built to plug into the next one just as easily.
Harness Commander runs underneath everything he does — a self-correcting, compounding intelligence harness, built in thirty days and running in production every day since.
The field named them. The edge here was wiring them into one loop that compounds — boot with the right context, work inside structured memory, close with judgment. The next session starts smarter than the last.
And the OS wasn't the only thing. Caby — an autonomous Telegram bot. A self-running video pipeline. All from the same instinct: stop hunting for projects, solve the real problem in front of you.
Four moments. Click any to see what's behind it.
Claude assessed the harness against a global benchmark of practitioners. No self-reporting. No external panel. The system was audited, scored, benchmarked — and the result surprised even Raj.
Among Claude practitioners worldwide. This intersection — CS domain depth + AI infrastructure builder — doesn't exist in market.
First-ever run of the system-design-check agent. Caught a regression and fixed it autonomously in the same run.
The long-term vision is products. But first — join an organisation serious about making AI work at scale, and sharpen harness engineering at enterprise level. Harness Commander is the proof of work. This site is the portfolio.
A role where I implement AI harness engineering at enterprise level — helping teams build reliable, repeatable AI workflows that actually ship
Organisations serious about the gap between AI capability and AI reliability in production
Teams where CS domain expertise + AI infrastructure thinking = genuine competitive advantage