Achint Mehta
Senior software development manager & AI security researcher
Achint Mehta is a senior software development manager and AI security researcher. His research looks at how large language models behave as autonomous coding agents — most recently an observational study of how reasoning effort, rather than tool access, drives first-try reliability in agentic code generation.
This site is where he writes about the things he builds and studies: how LLMs and AI agents work from the ground up, low-level systems and C/C++, networking, and the history of the programming languages we all stand on. Start with the AI Arena series, the blog, or the technical articles.
Publications
- The reach of a verification tool decides its value: A controlled study of verification surface, artifact quality, and cost in AI coding agents — arXiv preprint, 2026.
- Reasoning effort, not tool access, buys first-try reliability in agentic code generation: an observational study — arXiv preprint, 2026.
Contact
The best way to get in touch is by email: hello [at] achintmehta [dot] com