Work with us

Request a model review.

Audits, retainers and workshops for people who have to show that an AI system works. We respond within two working days.

Commissioned audits

A scoped review of a credit, risk or decisioning model: bias, stability, explainability, and a written finding. Credit-model work first.

Advisory retainers

Ongoing model-risk review for a bank, insurer or supervisor that wants a second pair of eyes on new systems.

Workshops

Short sessions for finance and risk teams on what an AI model audit actually checks.

What we will not do

Write a conclusion to order, let one client exceed 40% of revenue, or put our name on a result a second person has not reproduced. Independence

Request

Tell us what you need reviewed

Direct

hello@thesafelab.co

Book a 30-minute call

Join the Network if you want to contribute research, not commission an audit.