Fifteen years building AI systems that ship
I have built regulated AI systems myself, repeatedly, and learned where they break. Now I advise the organisations doing it.
Vikram Jha Advisory is an independent practice. Independent is a claim about neutrality, not about headcount: assurance work is only worth anything if the person doing it has no stake in the platform being assured. I hold none.
Delivery runs through me plus a standing network of senior engineers I have shipped with before. You are not buying a bench of juniors with a partner’s name on the cover — the person who scopes the work does the work.
Background
I led an international team in the DARPA Robotics Challenge, where humanoid autonomy had to work in field conditions that forgave nothing. The judgment formed there — about what autonomous systems do when reality disagrees with the plan — is the foundation of the governance work I do today.
Since then: production AI across financial services, insurance, healthcare and regulated industry, in the United States, India and Europe. Founder-side and advisory-side both.
Also currently
I am co-founder of an AI infrastructure company, in stealth, building the authorization layer that decides what an agent is permitted to do, enforces it at the point of action, and produces evidence a regulator will accept.
I do not discuss specifics publicly, and I disclose the relationship in any engagement where it could bear on my advice. If you are working on agent governance at an enterprise or as a research group, I am glad to go deeper under NDA.
How I work
- A small number of engagements at a time. Depth is the product.
- Written output your risk function can take to a committee, not a slide deck.
- I will tell you when the answer is that you do not need me.
- Remote, across United States, India and Europe hours.
If any of this is live for you
The fastest route is a conversation. Bring the architecture you are worried about — the first useful thing usually surfaces inside twenty minutes.