AI Engineering: your own AI infrastructure and AI agents for your business
We build an AI platform with you that you own and that runs in Switzerland, and we develop AI agents for marketing, sales, customer service and administration. Secure from the start, because security is our core business.
Engineering with security at the core
- Security is our core business: permissions, prompt injection and data leakage are part of the architecture from day one.
- Vendor-independent: we choose hosting and models for data protection, performance and cost, not for partner programmes.
- From practice: our own GRC agents handle ISMS routine work for clients. What we build for ourselves, we build for you.
An AI platform that belongs to you
Many companies want to use AI without handing customer data, contracts or source code to foreign cloud services. We build your own AI platform: with a Swiss cloud provider, in your data centre or a combination of both.
Hosting in Switzerland
Swiss cloud, your own data centre or both. We compare providers and recommend what fits your data and your budget.
Models of your choice
Open language models in your own environment or commercial models via protected interfaces. The right model for each use case.
Knowledge from your data
Documents, manuals and systems become usable for AI (RAG). Everyone sees only what they are allowed to see.
Access, logging, cost
Sign-in with your existing login (SSO), roles, logging of every request and cost control per team.
AI agents for your workflows
An agent completes a task from start to finish: it reads, decides according to your rules, works in your systems and submits the result for approval.
Marketing
Prepare content for website, LinkedIn and newsletters, optimise copy for search engines and AI search, evaluate campaigns.
Sales
Research and qualify leads, prepare quotes, keep the CRM up to date and remind you to follow up.
Customer service
Triage or answer requests, create tickets, answer from your knowledge base.
Administration and finance
Capture receipts and invoices, review contracts, compile reports.
HR
Pre-screen applications, support onboarding, answer internal questions about policies.
Compliance and security
Review evidence, assess suppliers, check policies. This is how our GRC agents work too.
We connect agents to your systems, such as CRM, ERP, ticketing, Microsoft 365 or Google Workspace. A person decides on every action with consequences.
If you work with Microsoft 365
Copilot & Agent 365
Introduce Copilot and AI agents securely: readiness, rules, approvals.
Learn moreData Security (Purview)
Find, label and protect sensitive data before AI can access it.
Learn morePrivacy & Risk Management
Implement data protection under the revFADP and GDPR in Microsoft 365.
Learn moreFor your home too
The same principles for flats and houses: smart switches without a bus system, a local hub without cloud dependency, a secure network and AI agents for energy, security and comfort.
Secure from the start
- Threats under ISO/IEC 27090 built into the architecture: prompt injection, data leakage, poisoned data
- Least privilege for every agent, human approval for critical actions
- Security testing before go-live, including attacks on the model
- Logs and evidence for audits, ISO/IEC 42001 and the EU AI Act
From idea to agent in production
Use case workshop
- Find the workflows with the greatest benefit
- Clarify data and systems
- Estimate effort and benefit
Prototype
- One use case with real data
- Test with future users
- Decide on the roll-out
Build
- Infrastructure and integration
- Security testing
- User training
Operation
- Monitoring and updates
- More agents as needed
- As a managed service if you wish
We bill projects by effort or at a fixed price per quote.
Frequently asked questions
Do we have to use a specific vendor such as Microsoft?
No. We are vendor-independent and recommend hosting and models based on your requirements for data protection, performance and cost. If you already work with Microsoft 365, Copilot and Agent 365 can be included.
Where is our data?
Entirely in Switzerland if you wish: with a Swiss cloud provider or in your own data centre. Together we define which data a model may see.
What does a custom AI agent cost?
It depends on the task and the connection to your systems. After the use case workshop you receive a quote with project effort and running costs.
Who runs the agents afterwards?
Your IT, us or both together. We take over monitoring, updates and further development as a managed service if you wish.
How do you prevent an agent from making mistakes or being misused?
Every agent gets only the permissions it needs. A person approves critical actions and everything is logged. Before go-live we test specifically against prompt injection and data leakage.
Have an idea for an AI agent?
We check with you whether the use case pays off and how to implement it securely.