Richik Pal

Hi! I’m Richik, a Computer Science and Molecular and Cellular Biology graduate from UC Berkeley. I’m building Naia, an AI company focused on Indic language data and LLMs, and I’m interested in making language technology work better for communities that are still underserved by today’s AI systems.

I work across agentic AI, language models, and neurobiology. At BAIR, I co-authored AgentTrace, a runtime observability framework for LLM agents accepted to the AAAI 2026 LaMAS Workshop. At AWS, I’ve worked on Amazon Q, including multi-agent query rewriting workflows with Strands SDK and MCP tools to improve retrieval quality for RAG systems. I’m also an incoming Software Engineer at Amazon Q.

Before that, I spent time in research and teaching: decoding surprisal from fMRI brain signals under Jack Gallant at the Gallant Lab, wet lab tardigrade transgenic experiments under Professor Saul Kato at UCSF, and course staff roles for Agentic AI and Algorithms. I’m especially excited by systems that sit between research and deployment: reliable agents, multilingual AI, human-computer interfaces, and ML tools that become useful in the real world.

Outside work, I love high-altitude travel and long outdoor days. I grew up in Mumbai, have trekked through the Himalayas, and have been lucky to explore glaciers in Greenland and volcanoes in Patagonia. I’m always happy to connect with people working on Indic AI, agentic systems, language tech, neurotechnology, or ambitious ideas at the edge of science and engineering.