Compliance and privacy first
We prioritize legal compliance and privacy at every step in the pipeline. Payroll, PII and commercial data stay out of scope by design, not as an afterthought.
We're hiring
Help unlock the data that trains real-world AI
Petrarch unlocks the specialized knowledge embedded in economically vital businesses to accelerate real-world AI. We are hiring a Founding Engineer in San Francisco.
Open roles
1 open roleYesterday's data serving tomorrow's economy
The world runs on information, and quality is getting harder to find. Open datasets and public repos are no longer enough to train state of the art models or build long-horizon benchmarks. The next frontier exists within the operational records of real businesses.
That work means going where the record actually lives: the Slack threads, Jira tickets, Confluence pages, Grafana alerts, Snowflake tables, GitHub pull requests and SAP works orders that a company's day is written into. We build long term partnerships with companies of all sizes to maximize long-term value and advance research insights.
We are backed by Y Combinator.
Partnerships, not extractions
We build long term partnerships with companies of all sizes, so the value compounds for both sides rather than ending at a single handoff.
Close to the real work
The knowledge we care about sits in operational records from the shop floor to dispatch. We go and understand that work before we model it.
Come build the foundation
If the specialized knowledge inside real businesses is the problem you want to work on, we would like to hear from you.
See open roles →