How do you handle GDPR and UK GDPR?

We design to comply with GDPR and UK GDPR, with a data processing agreement and logged access.

Private AI

Your data. Your rules.

Have documents that cannot leave the company? We design AI within your perimeter: cloud on European servers, a hybrid solution or models running on your own servers. With controlled access and a log of every operation.

Let's talk about your constraints Compare the options

The problem

You want AI, but not at the cost of your data.

Here are the most common concerns and how we address them. And there is much more.

ProblemContracts and customer data that cannot end up in the cloudSolutionModels running on your servers, with no external transfers ProblemYou don't know where your data goes when you use AISolutionDocumented architecture: you always know where it is and who sees it ProblemEach department must see only its own documentsSolutionPermissions by role, person and source, linked to your identity system ProblemYou need to prove who did whatSolutionA log of requests, answers, approvals and changes ProblemGDPR and UK GDPR requirements to meetSolutionDesigned to comply with the rules, with a data processing agreement And much more: we start from your constraints, not from the technology.Tell us yours

Deployment options

Cloud, hybrid or on your servers. We decide together.

It depends on data, constraints, volumes and the systems to connect. Click an option to see the differences.

Does it work offline too? A model running locally can work without internet, but some integrations need a connection. We check this case by case.

Governance

Who sees what. Who decides what.

Written rules, verifiable permissions and a log that answers auditors' questions.

Every person and every agent sees only what their role allows.
Restricted folders stay closed to anyone not authorised.
Important actions are confirmed by a person.

A log of requests, answers, approvals and changes.
Model versions tracked, with periodic evaluations.
Data retention defined by contract. Data is never used to train models.

Architecture and integrationsHow it connects to your systems with minimal permissions and approved writes.TEKO-AI, on your servers tooThe business AI assistant with role-based access and exportable data.

FAQ

Common questions. Straight answers.

Where does the data go?

To the cloud on European servers, a hybrid setup or your own servers: we decide together based on the data and the constraints.

Is our data used to train the models?

No, never.

Can we keep everything on our own servers?

Yes: we design on-premise solutions, with models running on your infrastructure.

Does it work offline too?

It depends on the deployment: a local model can work without internet, but some integrations need a connection. We check this case by case.

What if one day we no longer need it?

We switch it off. Data and memory remain yours and can be exported.

Read all the answers on costs, data and method

First step

Have data that can't leave? Let's start with the constraints.

Tell us about your infrastructure, regulations and systems: we will propose the best-suited option.

Book a call Architecture and technology