Dashboards and reports
Usage insights
See what kinds of work people use AI for, from category labels on sampled requests, without storing prompts.
What it is
A classifier labels sampled chat requests with a category from a built-in taxonomy of 22, plus your own categories and one level of subcategories. Reports show categories over time, each category's share and confidence, its subcategories, models and people, the mix per team, and the classifier's own latency, failures, tokens and cost.
Why you want it
Leadership asks what AI is actually used for before deciding which models, training or policies to invest in. Labels answer that from real traffic while the prompts themselves stay unstored.
How it works
- Built-in categories such as Coding, Technical help, Legal, Medical and health, and Writing and editing; built-ins can be renamed or disabled
- Any decision model can classify, or any chat model prompted to answer in JSON
- Background mode adds no latency; inline mode returns the label in X-Janus-Category within a time budget
- A classifier failure never fails the request
- Reads the latest message with the recent conversation as context, and classifies what the caller wrote, not instructions Janus adds
- Category and subcategory are asked as one question when the classifier accepts enough choices
- Try it classifies a sample with unsaved settings to compare classifiers and speed
- Only labels, confidence and timing are stored; results kept 180 days by default
POST /v1/chat/completions
{ "model": "chat-default", … }
HTTP/1.1 200 OK
X-Janus-Category: coding/rust
See it on your own network.
The Community edition is free for up to 25 people. The 30-day Business trial unlocks every Business feature.