What it does
A Python library for checking an agent's work with several Jev questions in one call.
Benchmarks & research
jevals
Bookmark: jevals
Keep this in your collection.
Leave a noteWhat would you try with this?
: jevals
Only you can see your notes.
Screenshot unavailable. Open the experiment ↗
How you can use it
You can use this to automatically check an AI chatbot. Imagine your app searches a store catalog. This tool can verify if the bot actually used those search results. A developer sends the chatbot's messages to TypeSafe. TypeSafe is an outside AI service. It answers several questions about the chat at once.
Your developer needs an access key that connects the app to TypeSafe. The tool limits how much history it checks. It reviews the last eight messages of a conversation. It also looks at up to twelve tool actions, like search results. It shortens very long text so everything fits.