Frequently asked questions
Short answers. If you want to read on, every topic has a page or a knowledge base article.
What it is
What is Plainwater AI in one sentence?
A workspace where your team works on data with any AI model, from question to dashboard, while a gateway decides what the model gets to see.
Is it a chatbot on my database?
No. A chatbot often sends rows to a model. Plainwater AI starts from metadata, lets rows through only when you allow it per source, and makes names pseudonymous.
Does it replace my warehouse or ETL tool?
No. It works on your existing sources, warehouse and reports. With Yres it works together through the Yres MCP server.
Data and security
Which data goes to the AI model?
You choose per source: metrics only (no rows), encrypted values (every cell a token) or open (literal values, only on sources you allow). The strictest rule always wins.
Does the model see my table names?
No, the model sees pseudonyms such as t_0056. The platform translates the answer back to the real names for you.
Does the AI get my database password?
Never. The agent gets no model provider key and no customer credential, only a one-time bootstrap code. Customer secrets are encrypted per source.
Can I see afterwards what the AI did?
Yes. Every step is in an audit chain that shows gaps, and prompts and answers are stored encrypted. Whoever reads them is logged too.
Models and tools
Which models can I use?
Claude, GPT, Gemini and Mistral are set up, and any model with an OpenAI-compatible API is configuration. Per customer you choose which models are allowed.
Does it work with my own coding AI?
Yes, through MCP. Claude Code, Codex, GitHub Copilot, Cursor and other MCP clients can connect. The rules apply there too.
Do I pay for model usage when I use my own subscription?
Not in the Solo way of working: everything runs in your own AI tool on your own subscription, and the platform makes no model costs.
Europe and hosting
Where does the platform run?
Today in Azure West Europe (the Netherlands). We are moving to a European cloud with a European owner, such as StackIT or Hetzner.
Can I enforce EU models only?
Yes. Per customer you set which models are allowed, for example only Mistral and Azure OpenAI with EU data residency.
See it on your own data
In 30 minutes we show the platform on a sample environment and discuss what it would see in yours. After that you know whether it fits.