Private AI

The benefits of AI, while your data stays inside your company.

Price
Setup from EUR [PRICE] + hosting
Tools I typically use
Ollama · Open WebUI · Swiss / EU hosting · open-source models

AI for a practice that can't share patient data

Before

Staff would like AI help with letters and summaries, but public AI tools are off-limits for patient information.

After

A private AI runs on a dedicated server. Staff draft and summarize documents, and no data leaves the practice.

Your own AI, under your control.

  1. 01

    Self-hosted models

    Open AI models running on your own server or a Swiss or EU host.

  2. 02

    Private team chat

    A familiar chat interface for your staff, with individual user accounts.

  3. 03

    Ask your documents

    Questions about internal files answered without uploading them to a public cloud.

  4. 04

    Access control

    Decide who can use what, with activity logs.

  5. 05

    Privacy by design

    Set up with the Swiss data protection act and GDPR in mind.

  6. 06

    Maintenance

    Models and software kept up to date for you.

Who it's for

  • Medical practices
  • Law firms
  • Fiduciaries & accountants
  • HR teams
  • Consultancies
  • Any business with confidential data

Common questions

Is it as good as ChatGPT?

For everyday tasks like drafting, summarizing and answering questions about your documents, open models are very capable. For some advanced tasks cloud models are stronger, and I'll be honest about the trade-off.

What hardware do I need?

Either a server in your office or a rented server. I recommend a setup that fits your team size and budget.

Can it be combined with cloud AI?

Yes. Sensitive data stays private, and less sensitive tasks can use cloud models if you want.

Want AI without the privacy worries?

Tell me briefly about your business. I reply within one working day with times for a free 30-minute call, then send a written fixed-price quote.

  1. 01 Talk. A free call about your situation.
  2. 02 Quote. Fixed price, scope and timeline in writing.
  3. 03 Build and test. Set up with your real data.
  4. 04 Support. Adjustments when you need them.