Deepak Joshi — full-stack, mobile and ai/llm engineer

I started at Deloitte USI, which is where I learned what "production" actually costs — the review gates, the audit trails, the fact that a change nobody can trace is a change nobody can defend. That experience is the reason a CI pipeline, a security review and a handover document are part of every project here rather than something you pay extra for.

I am also the founder of Dharmarthlabs, which means I have been on the other side of the table: paying for engineering, waiting on a delivery, and discovering exactly which shortcuts cost more later. It changed how I scope work. Discovery happens before a price, the price is fixed against a written scope, and changes get quoted instead of quietly absorbed.

Code Hippies is the studio I build under. The work runs from regional Hindi news platforms serving Uttarakhand, Himachal and Haryana, to a consultation funnel for a private dietitian, to a 21-kilobyte static site for an ashram that cannot carry a maintenance contract. Different stacks, because those were different problems — the technology choice follows the constraint rather than a preference.

The current focus is AI and LLM engineering: retrieval-grounded assistants that answer from documents a business controls, with an evaluation harness that runs in CI and includes questions the system is supposed to refuse. The assistant on this site is built exactly that way, and you are welcome to try to break it.