TrueFetch + Jev by TypeSafe
From web data.
To your next decision.
Turn job listings into a focused shortlist. Jev evaluates the text; your code decides what moves forward.
Job records
Bring a JSON file with titles, descriptions and original source links.
Your dataJev
Check role fit, work arrangement, evidence and instructions hidden in the text.
One call per listingYour shortlist
Keep likely matches, skip clear mismatches and send uncertainty for review.
Rules you controlStart with your own job data today. TrueFetch collection is in development; the downloadable workflow calls Jev directly using your TypeSafe account.
A decision you can inspect
Keep the signal.
Keep the source.
Every evaluated record includes the model version, probabilities, token usage and original URL. Change the routing thresholds in one place.
See the routing rules →Clear role match and the requested work arrangement.
Missing duties, uncertain answers or suspicious instructions.
Clear mismatch with your role or work arrangement.
Start building
A small workflow.
A useful starting point.
Python 3.11+ for the complete filter. Node 24+ for a minimal TypeScript API call. Both run outside the browser.
Independent TrueFetch workflow using the TypeSafe API. Jev is a TypeSafe model. Explore the Job Scraper API →