We use a few strictly-necessary cookies, no marketing trackers, and cookieless analytics. Read our cookie policy, or pick a preference.
We're a small Manchester team building the system of record for AI obligations. It's a genuinely hard problem at the seam of law, machine learning, and engineering, and we're looking for people who want to work at that seam.
AI is now governed by hundreds of instruments across dozens of jurisdictions, and the count rises every quarter. The work here is to turn that sprawl into a live, cited, machine-readable record an enterprise can actually operate against. It's real, unglamorous depth: reading consolidated text, decomposing obligations, and building systems that produce evidence rather than promises. If that sounds like the good part, we should talk.
Lawyers and ML engineers work the same problem side by side. Neither throws it over a wall to the other.
150,000-word guides, thousands of atomic obligations, every one cited. We go all the way down and get it right.
We build things that run and produce evidence, not decks and demos. The output is a working system.
Manchester-based, deliberately lean. Everyone owns real surface area, and the work you do is visible.
These are the disciplines we hire into. Roles flex to the person; if your strength spans a couple of these, say so.
Turn regulation into structured, cited obligations. For lawyers who can think in systems, or engineers who can read a statute.
Retrieval, decomposition, and evaluation over a dense regulatory corpus. Build the graph that makes the record queryable and trustworthy.
Make the record live: the platform, the bindings to real systems, and the tamper-evident spine underneath.
Embed with customers on their hardest cases: contested classifications, regulator conversations, delivery on a deadline.
Make dense, high-stakes regulatory information legible and calm. Product and brand design in a serious register.
Bring the platform to regulated enterprises: partnerships, design-partner relationships, and the field motion.