Trust Center
What we do,
stated plainly
How we work with your systems
Seven practices that hold on every engagement. They are things within our control, which is why we are willing to write them down.
Least-privilege access
Access is requested per engagement, scoped to the systems the work actually touches, and revoked at handover. We do not keep standing credentials to a client environment after a project ends.
Your data stays in your systems
Wherever the architecture allows it, data stays in your own accounts, databases and tenancies. We build in your environment rather than copying your records into ours.
Secrets never go in source control
API keys, tokens and connection strings live in a secret store or environment configuration you own. Repositories carry references, never values.
Audit trails agreed in writing
Before any data is ingested we agree what gets logged, where logs live, who can read them and how long they are kept. Retention is a decision made up front, not a default nobody chose.
Subprocessors disclosed
If a build depends on a third-party model provider, host or service that will see your data, we name it before the build starts so you can accept or reject it.
Human review on consequential actions
Anything that sends money, contacts a customer, changes a record of truth or makes a decision a person would be accountable for goes through an explicit review gate. Autonomy is something you grant deliberately, step by step.
You own the code and the credentials
Source, infrastructure definitions, documentation and accounts transfer to you. Nothing we build is designed to require our continued involvement to keep running.
Need this against your own checklist?
Send us the requirement rather than a questionnaire template. We will answer each point directly, including the points where the answer is no.
Send us the requirementWhere regulation fits
The honest division of responsibility, so nobody discovers it late.
You know which rules apply to your business. We do not, and we will not pretend to. What we do is take the requirements you give us and turn them into constraints the system has to satisfy: where data may be stored, which fields must never leave your environment, what has to be logged, what a person has to approve before it happens.
Those constraints get written down before the build, implemented in the architecture, and demonstrated at handover. If a requirement cannot be met with the tools in scope, we say that during scoping rather than after delivery.
We are not a compliance authority, an auditor or an advisor on your obligations. The determination that a system satisfies a regulation is made by you and the people you take that advice from. Our contribution is engineering that makes the determination possible to reach.
Our own terms and data handling are set out in writing. Read the privacy policy and the terms.
Boundaries
What we are not
A trust page that only lists strengths is not a trust page. These are the things we cannot offer, stated before you ask.
We are not certified
We hold no SOC 2 report, no ISO 27001 certificate and no third-party security attestation, and we will not describe ourselves as working toward one until that is a fact we can point at. If a certified vendor is a requirement for your procurement process, we are not that vendor.
We are not a compliance authority
We cannot tell you that a system satisfies a regulatory obligation. Whether your use of a system meets the rules that apply to you is a determination for you and your legal or compliance advisors, not for us.
We do not issue assurance
No seals, no badges, no audit letters, no questionnaire answered on anyone else's behalf. What we can give you is a plain description of how a specific build handles access, data and failure.
We are not a managed security provider
We are an engineering team. We are not a round-the-clock monitoring service and we do not take over operational responsibility for your estate unless that is scoped and agreed as its own arrangement.
Reporting a security concern
If you believe you have found a vulnerability in something we built, or you have a question about how an engagement handles your data, use the contact form and mark the message as a security issue.
We will acknowledge it, tell you what we are doing, and come back to you with the outcome. We do not run a bug bounty programme and we will not pretend to operate a 24-hour response window we cannot staff.
Trust questions, answered directly
Including the ones where the answer is no.
No. We hold no security certification or third-party attestation. We are a small engineering consultancy, and rather than imply otherwise we describe the practices we actually follow on an engagement and let you judge them. If certification is a hard requirement for your procurement, we will tell you plainly that we do not meet it.
We can build toward requirements you define. If your obligations mean data must stay in a particular jurisdiction, that certain fields are never sent to a third-party model, that every access is logged, or that a human approves a given step, those become architectural constraints we design and implement against. What we cannot do is certify that the result satisfies the regulation. That assessment is yours.
In your systems wherever the architecture allows. The default is to build inside your accounts and tenancies, so the data never leaves infrastructure you control. Where a build genuinely needs data somewhere else, we identify it, name the service that will hold it, and get that agreed before anything is moved.
It depends entirely on the build, which is why we disclose it per engagement rather than publishing a generic list here. Any model provider, host or service that would process your data is named before implementation starts, and you can rule any of them out.
It is revoked. Handover includes removing our accounts, rotating any shared credentials and confirming in writing which access has been withdrawn. Ongoing maintenance, if you want it, is a separate scoped arrangement with its own access.
Use the contact form and mark the message as a security issue. We will acknowledge it and tell you what we are doing about it. We would rather hear about a problem early than find out later.
Have a requirement you need answered?
Send us the specific constraint rather than a generic questionnaire. We will tell you how we would build against it, or that we cannot.