I was a post-doctoral researcher at Texas A&M University working on large language models. I am update my personal research at www.llmnerd.dev. For many iterations, I assisted TAMU ECEN flagship ML course: ECEN 740 taught by Prof. P R Kumar.

I consider myself luckiest to have met my Doctoral Advisor, Dr. P. R. Kumar.  He introduced me to the beautiful art of solving the most challenging systems problems in a minimalist way.  I am equally excited and grateful to have defended my thesis with Dr. Paulraj on my committee.

I have dedicated quite a lot time developing deep understanding of machine learning, LLM post-training and controlling agents. 

These days, I mostly work on LLM fine-tuning, and developing agents. I worked on Web Agents at Amazon. Now, mostly focused on Coding Agents, and Voice Agents.

Tutorvi: Voice only conversational tutor, that can retrieve relevant information to it context with <9ms latency.

I completely rebuilt an Agentic IDE with modern features like efficient context compression, caching, LSP, and defacto deep we research.

In the past, I developed Cellular Standards, built IoT wireless networks, and solved some challenging problems in directional networks.