Genomi gives AI agents a local database for personal genome questions
The open-source project parses raw DNA files into a local database instead of sending genetic data into a model context window.
By Ryan Merket · · updated
Why it matters
Genomi sits at the collision point between local AI agents and consumer genetic data, where privacy controls and evidentiary limits matter as much as model capability.
Hongjian Zhou (@itsEmZee_) introduced Genomi, an open-source harness meant to let AI agents answer questions about a user's DNA, in a thread on X on Tuesday.
https://x.com/itsemzee_/status/2061795034840629329?s=46
Zhou framed Genomi around a familiar consumer-genomics problem: people take a DNA test, scan the report, then leave the raw file unused. He wrote that he recently gave his own DNA data to a coding agent and saw a need for a safer interface between personal genomes and general-purpose AI.
Genomi, according to Zhou, parses a raw DNA file into a local database that an agent can query, rather than pasting the file into an AI context window. "Your raw DNA file should not be dumped into an AI context window," Zhou wrote, calling it "too big, too sensitive, and too easy to misread."
The project is still early, Zhou said. He claims Genomi gives an agent 88 domain-specific tools across 30 evidence categories and is designed to answer whether a variant was measured, whether data supports a question, or when the system should say "No" or "I don't know." He also said the raw DNA file stays on the user's machine through a local database the team calls Active Genome Index, with an update command intended to sync public genetics research over time.
Zhou said Genomi was built with MT (@MatthewZMD), Jinge WU (@JingeWok), and Andrew Liu, and can work with raw DNA files from 23andMe, AncestryDNA, sequencing providers, Dante Labs, and others. The unresolved question is how users should validate agent outputs in a field where a plausible answer can still be medically consequential.