Skip to content
HMD Digital
Trust

Your data, your process, our accountability.

HMD Digital builds and manages AI agents for SMEs, and stays accountable for them in production. Here is where they can live, how your data is handled and how your people stay in control.

Start a discovery conversationAsk for our DPA
GDPREU processing by default
DPA availableWith every client
EU hosting optionOr your own cloud
Human oversightDesigned in, not added
Where agents are deployed

Where it fits your business.

Our infrastructure, your cloud, a hybrid setup or on-premise. The choice is yours, and it goes into the proposal before you decide.

Our infrastructure

Where things live

The simplest start. We host and manage everything, with EU processing by default.

Data handling and GDPR

Only the data the process needs, handled in the EU.

EU processing by default

Unless we agree otherwise with you, your data is processed in the EU.

Only what the process needs

Agents get access to the data the process uses and nothing more. What they do not need, they never see.

A DPA with every client

Every engagement comes with a data processing agreement that sets out what we process, where and why.

Our EU AI Act stance

Oversight, transparency and logging, by design.

We build every agent so that people can see what it does, step in when it matters and trace any output back to its source.

Human oversightPeople can see what an agent is doing, step in and override it. Decisions that need judgment go to a person.
TransparencyYour team knows where agents are at work, what they handle and where their output comes from.
LoggingAgent steps and human decisions are logged, so any output can be traced back and reviewed.
Who owns what

Clear lines, agreed before anything goes live.

Yours

  • Your data and documents
  • Everything the agents produce for you
  • Your process knowledge
  • The final say on decisions

Ours to answer for

  • How the agents behave in production
  • Monitoring, fixes and improvements
  • Keeping agents current as your process changes
  • Telling you when something goes wrong

Agreed in the proposal

  • Where the agents are hosted
  • Who owns the infrastructure
  • Access, roles and approvals
  • How value is measured
Human in the loop

Agents prepare. People decide where judgment matters.

Approvals and edge cases go to your people. Where each human step sits is designed with the team that does the work.

  1. Work arrivesA request, a document, an email
  2. Agent preparesReads, checks and drafts
  3. Exceptions flaggedAnything unusual is marked
  4. PersonA person decidesApproves, edits or rejects
  5. Done and loggedWith the sources attached
Monitoring and incident response

If something goes wrong, we feel it before you do.

Every agent in production is watched for performance, costs and exceptions. When something is off, this is what happens.

DetectPerformance, costs and exceptions are watched in production, so we notice first.
ContainPause the agent or route the work back to your people, so nothing wrong goes out.
FixFind the cause and fix it, with the control we kept for exactly this.
Tell youYou hear what happened, what we changed and what it means for your process.

Questions about your data? Ask them on the first call.

We'll walk through where your agents would live, what they would see and who stays in control.

Start a discovery conversation