TofuPilot AI
Last updated on May 21, 2026
TofuPilot AI turns your test data into answers and actions. Chat queries every run, unit, and measurement in plain English; the agent runs a guided root cause analysis on demand; alerts watch the production line and page the right people when something breaks.
Chat
Ask questions across runs, units, parts, and measurements.
Agent
Run an autonomous root cause analysis on failing units.
Alerts
Detect yield drops, drift, and failing reference samples.
AI provider
TofuPilot AI runs on open-source large language models hosted by our partner Infomaniak in Switzerland. Every prompt, tool call, and response stays inside the European Union and the Swiss legal framework. Your data is not shared with OpenAI, Anthropic, Google, or any third-party model provider, and it is never used to train a foundation model.
| Property | Value |
|---|---|
| Model provider | Infomaniak AI Tools |
| Hosting region | Switzerland |
| Data residency | EU + Swiss legal framework |
| Training opt-in | No. Your prompts and data are never used to train models. |
| Sub-processors | Infomaniak only. No US-based AI provider in the chain. |
If your compliance team requires a different setup, the self-hosted deployment lets you point TofuPilot AI at any OpenAI-compatible endpoint you operate yourself.
How is this guide?
Setup Tokens
Learn how TofuPilot setup tokens pre-authorize the CLI as a station with one-hour, single-use credentials redeemed on the test machine.
AgentPreview
Run an autonomous root cause analysis on failing units with the TofuPilot Agent: hypothesis generation, evidence collection, and statistical scoring inside the chat.