Automating
Put AI at the centre of your compliance function.
We use Claude to build compliance tooling inside your firm — policy libraries, surveillance specifications, monitoring programmes, risk and control frameworks, SMCR maps and horizon tools. All linked, all yours, no licensing cost.
- Who it’s for
- CCOs and compliance leads who want the tooling they’ve been imagining, without a vendor project
- What you need
- An enterprise AI licence and a willingness to feed it your firm’s context
- How we engage
- Introductory ten-day engagements, then a build-out at your pace
What we build
Policy and procedure library
Harmonised, searchable and updated en masse when rules change.
Monitoring and testing
Plans generated from your risk profile, with outcomes captured and reported.
Risk and control assessment
Each risk tied to its control, policy, rule, activity and owner — and kept current.
Regulatory horizon
Rule changes and final notices turned into firm-specific impact assessments.
Regulatory operations
Surveillance, best execution and reporting workflows: ingestion, reporting and record keeping.
Training and trackers
Training built around your people and perimeter, and trackers for work in flight.
How it works
Step 1
Your licence
We work inside your firm’s AI environment and data governance, never ours.
Step 2
Your context
We feed it your permissions, activities, policies and procedures.
Step 3
Build
Working tools, built and tested with your team in days.
Step 4
Hand over
Your team learns to work with it daily, then automate procedures with agents.
We don’t license software. We build it for you.
Compliance software is built for the average firm, so it fits no firm exactly. The gap becomes manual work for your team.
A tool built from your permissions, activities and policies fits from the start, and can be changed the same afternoon you decide it should be.
James builds his own compliance tooling this way. It is compliance expertise applied through AI, rather than a generic AI consultancy.
Questions firms ask
Do we need to buy software?
No. Everything is built inside your firm’s own AI environment, using your enterprise licence — Anthropic’s Claude in our case. There is no seat licence and no annual renewal, and the tools keep working if we stop working together.
What does the AI actually do without supervision?
Nothing that needs a compliance judgement. AI builds, drafts, monitors and summarises; a person with regulatory expertise reviews the output and makes the calls. We also help you set the data governance boundaries before any firm data goes in.
How long before we see something useful?
Days, not quarters. Introductory engagements run for ten working days and are designed to leave your team with a working tool, not a slide deck.
Want to see it on your own firm?
Ten working days, remote, on-site or hybrid, designed to leave you with something your team will use.