Jay Shah -- AI Systems, Foundational Models & Research

AI Systems, Research & Engineering

I am a Senior Data Scientist at 6sense building production AI applications, foundational model systems, custom embeddings, and agent evaluation frameworks. On this site, I write in-depth technical essays documenting what I build, break, and learn across large language models, retrieval-augmented generation (RAG), statistical text watermarking, and autonomous developer agents.

10 writingsRSSllms.txtOpenAPIAPI Docs

Recent Research & Technical Articles

View archives →

Latest

How does a text watermark work?

A first-principles investigation of how generation-time text watermarks work, from a weighted coin to Gemma, and where the evidence stops.

ai-safety llm watermarking gen-ai research

August 18, 2026•50 min read

Read article

Can Dense Retrieval Beat BM25 for Entity Resolution? (And At What Cost?)

Testing if applying Matryoshka Representation Learning (MRL) to tabular entity data could bridge the cost gap, compressing embeddings enough to remain operational while still beating BM25 on corrupted queries.

machine learning search infrastructure embedding models +3

March 9, 2026•14 min read

Read article

Can Dense Retrieval Beat BM25 for Entity Resolution? (And At What Cost?)

File-Based Memory Is a Terrible Idea That Somehow Works

A production diary of tiered episodic memory in AI agents. Three markdown files, an SQLite database, and a lobster that somehow remembers what you said three days ago.

ai-agents openclaw memory production

February 28, 2026•15 min read

Read article

File-Based Memory Is a Terrible Idea That Somehow Works

Stop Renting Your Workflow: Building a Custom AI Coding Agent with Pi

Four tools, a settings file, and full control - how I built my own AI coding agent with Pi instead of paying $200/month for a CLI that keeps changing.

ai-agents coding-tools llm developer-tools +1

February 11, 2026•11 min read

Read article