What I build for you
- AI assistant on your own data. An assistant connected to your CRM (HubSpot, Salesforce), ERP (Exact, Odoo, SAP) or documents (Google Workspace, Microsoft 365). Your team asks in plain language and gets answers from your own sources.
- Chatbot on your handbook or policies. Employees ask questions about procedures, HR or finance. GDPR-compliant and hosted in the EU.
- Document checks. Automatic checks on reports, quotes or contracts. If something is missing or off, you get an alert.
- Customer service automation. On top of your ticket system. The AI sorts questions, suggests answers and handles simple questions itself.
- Voice agent. An AI that takes incoming phone calls, answers questions and records appointments or callback requests.
- MCP server for your systems. So your team works directly in your own tools through Claude or ChatGPT, with the permissions you decide.
- Smart analyses. Daily alerts on notable trends in your data, with an explanation of why.
Two ways to use AI
- In the background. An email, invoice or file comes in. The AI processes it and puts the result in your systems. Nobody has to do anything.
- On demand. An employee asks an assistant a question, or asks through Claude or ChatGPT. The AI finds the answer or carries out the action in your systems.
A combination often works best. Read more about AI in the background under Process automation
Honest advice here too. Not every problem needs AI. Sometimes a simple integration or a good search function is enough. Then I'll tell you.
How I work
- Introduction. We discuss which questions and tasks take the most time now. Free and without obligation.
- Analysis and proposal. I map which data and systems the AI needs and what it can and can’t do. You get a concrete proposal with a price.
- Build and test. I build the solution and test it with real questions and documents from your daily work.
- Go-live and management. I keep an eye on the quality of the answers. If your data or the AI model changes, I adjust it.
What does an AI assistant cost?
You pay a fixed project price. A simple chatbot on your documents sits at the lower end. An assistant that works across several systems or takes actions costs more. After that you can choose monthly management and continuous improvement.
You pay the costs of the AI model itself directly to the provider. They’re listed separately in the proposal.
After the introductory call you’ll get a clear price estimate. No surprises. See all pricing
Why Tovy
- One point of contact. You work directly with the engineer who builds and manages your AI solution.
- Process first, then technology. With a background in supply chain management, I first look at how the work runs now, and only then where AI helps.
- Data engineering as the foundation. An AI assistant is only as good as the data underneath it. I build the integrations and data pipelines myself. Data warehouse and reporting
- I use it myself. I build and use AI agents every day for my own business, for bookkeeping, social media and data analysis.
- Documented and transferable. Everything runs in your environment and is documented. You stay in control.
