AI agents for business.
We build AI agents that read mail and documents, answer requests, and work inside your CRM, ERP, and PIM. AI integration here means connected to your data, measured against an evaluation set, gated by human review, and handed over to run in production.
Live demo
Watch an AI agent at work.
A supplier invoice arrives by email. The agent does the rest: through the interfaces of the ERP and PIM, and where there is none, it clicks through the program like a person.
Sample with invented data. This is how the agents we build into ERP, PIM, and CRM work.
This workflow for your companyHow the demo runs: a supplier invoice arrives by email. The AI agent reads the amount, due date, and items from the PDF. In the ERP it books the invoice through the interface, without a single click. In the PIM it updates the purchase price, also through the interface. The tax adviser's portal has no interface: the agent opens it, uploads the receipt, and fills in the fields like a person. Before submitting, it waits for the team's approval.
What we build with.
- Models
- Claude, GPT, or open-weight models for EU and on-prem operation
- Methods
- RAG, tool use, MCP servers, evaluation sets, guardrails
- Connected to
- CRM, ERP, PIM, shared inboxes, document storage, phone
- In production
- Logs, cost per request, approval for risky steps
How we start.
-
Write to us
Two or three sentences about the workflow. We reply personally and suggest a short call if it helps.
-
Proposal
A fixed-scope plan with timeline, price, deliverables, and open assumptions.
-
Build & handover
We run the build, ship the system, and hand over the operating model.
Pricing and deliverables.
The scope below covers a production AI loop, not a demo: data access, model behavior, evaluation, deployment, and ownership.
Pricing
€4.5k to €45k
- Timeline
- 2 to 10 weeks
- Scope
- Fixed scope after the first technical scoping call.
What's included
- AI architecture and use-case scoping
- RAG or agent workflow connected to source systems
- Evaluation set, guardrails, and human review flow
- Production deployment with telemetry and cost visibility
- Operator and developer handover
Add-ons
- Evaluation harness €1.5k
- Voice or chat channel from €4.5k
- EU or on-prem inference path from €3k
Questions on cost, time, and delivery.
What does an AI agent cost?
€4.5k to €45k, as a fixed price after the first technical call. One workflow with one connected system sits at the lower end; several systems, a dedicated evaluation harness, and a voice or chat channel sit higher. Model usage comes on top, and the deployment shows it per request.
How long does it take to build an AI agent?
2 to 10 weeks to production, depending on how many systems the agent works in.
Is an AI agent GDPR-compliant?
We fix the data flow before the first model call: which data the model sees, where it runs, and what gets stored. Data that must not leave the EU takes the EU or on-prem path (from €3k). The legal sign-off stays with your data protection officer.
Which AI model do you use?
The one that solves the task and fits your data: Claude, GPT, or an open-weight model on your own infrastructure. We measure it against an evaluation set built from your real cases before we commit.
What can an AI agent not do?
It does not decide alone about money, contracts, or customers. Risky steps go through human approval, and whatever the agent cannot classify with confidence goes to a person. It also cannot find data that is stored nowhere.
Do you offer AI consulting without the build?
Yes, at a day rate: use-case assessment, architecture, and an implementation plan with a price. The build then runs at a fixed price, by the same people.
Try it on your own data.
Free, no signup. Each one is a slice of this service, scoped down to a single job.
