Slurf · Your expertise, extended.

Everybody has AI. Nobody has your method. Het verlengstuk van je expertise.

Slurf turns the way you work into an AI process your team uses every day, that you can offer to your clients, and that always stays yours. Your method. Run by AI. Signed off by you.

Your method, executable

No loose prompts, but your method: the steps, the checks and the moments that need an expert. Slurf runs it for every case, even when a case takes days. Need a sign-off or missing a file? Then only that step waits, and everything is recorded.

The right model for every step

Not every step needs the same model. Slurf handles that per step: frontier models in European clouds, open models on Slurf's own EU GPUs, and soon a local runner at your site. You decide what is sensitive, Slurf thinks along. And no model is ever trained on your files: nobody gets smarter from your knowledge.

Your expertise as a product

Clients use your method from Claude, ChatGPT or Copilot and get the result, never how it works. Your expert signs where the method requires it.

Purely European, at home in Microsoft

Slurf runs purely European, on Scaleway, on Kubernetes infrastructure we operate ourselves. Sign in with Microsoft Entra ID, with single sign-on for your organisation, read files from SharePoint with each person's own permissions and start cases from Copilot. Slurf can also run in your own Azure subscription.

The right screen for the work

Findings next to their source, tables you can correct, drawings you can zoom. Plug in your own screen as an MCP App; it shows in Slurf and in Claude or ChatGPT.

Claude writes. Slurf checks.

Claude and Copilot do what the person at the keyboard asks. Slurf works for your method: a required check cannot be skipped, whatever AI or a colleague drafted is checked independently against your standards, checklist and house style, and where a person must decide the case waits for an expert who signs off. Every case has its own chat, and the same case works from Claude, ChatGPT or Copilot over MCP.

Yours. Really.

Your voice, your templates, your examples. Built on open standards: Open Workflow, Agent Skills, MCP and MCP Apps. Export your method any time.

How a case runs

1. Start from ChatGPT, Claude, Copilot or Slurf.
2. Slurf follows your steps with AI and your tools.
3. Results appear in the right screen, linked to their source.
4. Your expert checks the findings and corrects them.
5. Signed by name, delivered in your template, with its trail.

What you can build

Contract due diligence for legal teams, year-end work for accountants, damage assessments for insurers, and a privacy consultancy's DPIA method sold to municipalities. Illustrative examples, not customer references.

Pricing

• Team (coming soon): €59 per organisation per month (€49 billed yearly), plus AI usage. Chat and documents with your own sources, your own workflows, skills and templates, SharePoint and Google Drive, hosted in the EU.
• Custom: one-off setup plus subscription. We build your method together with you, custom tools included, with expert sign-off and a recorded trail.
• Provider: offer your method to your clients from Claude and ChatGPT. You set the price; Slurf keeps 8% of your revenue through Slurf.

All prices exclude VAT.

Questions

How is Slurf different from ChatGPT or Claude?

Those are smart assistants: powerful, but they improvise. Slurf runs your method, in fixed steps, with an expert who signs off and a trail you can show afterwards. Under the hood, Slurf uses those models.

Can our clients see how our method works?

No. They see what they can request and what comes out. Your instructions, rules, examples and tools stay with you.

What if the AI makes a mistake?

That is why your expert signs off. They review structured data, not walls of text, and correct where needed.

Slurf is a product of Artificial Heroes · KvK 97298344 · hello@artificialheroes.com · Privacy policy · Sign in