What you can hold us to.

Your data

We take only the data a project needs, hold it in your environment or ours under a data-processing agreement, encrypt it in transit and at rest, and delete it on request or at the end of the engagement. Nothing you give us is used to train anyone else's model.

Your code and secrets

Code lives in your repository from the first week. Secrets live in a vault, never in chats, documents or code. Access is per person and removed the day someone leaves the project.

Intellectual property

You own what we build for you: code, models, designs and documents, assigned in the contract. We keep only our reusable tooling, which is listed in the brief.

How we use AI tools

We use AI tools in our own work for drafting, review and testing, under our own accounts, and we never send your confidential data to a tool outside the agreement. Every agent we build says what it is where the law requires, logs its actions, and hands over to a person.

The laws we work under

GDPR for European clients, CCPA for California, DPDP for India, PDPL for the UAE and Saudi Arabia, and the consent and disclosure rules of each country an agent operates in.

Security practice

Two-factor sign-in everywhere; dependency and vulnerability scanning on every build; backups verified monthly; incident response with a named owner and a 24-hour notification commitment.

Published numbers

From the first full month after launch we publish here, monthly: the median time to first reply on an enquiry, the share of stages delivered on the written date, and the uptime of what we run. The zeros are published too.