AI matching and semantic job search
Use pre-generated embeddings and vector search to power recommendation and matching pipelines.
How it works with the jobdata API
- Every job post comes with a pre-generated 768-dimension text embedding (
embed_3sh=true), so there is no need to embed millions of descriptions yourself. - Vector search ranks jobs by semantic similarity to a text such as a CV summary or candidate profile (
vec_text). - Combine semantic ranking with regular filters like country, work mode or experience level.
- Markdown descriptions (
description_md=true) are ready to use in LLM prompts and RAG pipelines.
Example request
curl -X GET -H "Authorization: Api-Key YOUR_API_KEY" "https://jobdataapi.com/api/jobs/?vec_text=senior%20backend%20engineer%20python%20postgres&country_code=DE&max_age=30"All endpoints are public with live data for testing/structure review in a limited fashion. Add an API key to remove anonymous limits according to your plan.
Recommended plan: access pro+
access pro+ includes embeddings and 2,500 vector search queries per day, access ultra raises that to 5,000. All plans are flat-rate with unlimited usage - no credits, no usage caps, no overage surprises.
Related docs and tutorials
Introduction to Using Vector Search and Embeddings through the jobdata API
Unlocking semantic job search and candidate matching by combining pre-generated embeddings with scalable vector databases.
Vector Embeddings and Search API Documentation
Enhance job searches with semantic understanding through vector embeddings.
Using the jobdata API for Machine Learning with Cleaned Job Descriptions
Build a Markdown-based pipeline for exploratory job clustering and skill extraction while preserving meaningful technical terms.
More use cases
Explore jobdata API with AI tools
Ask an AI assistant how to search and build with live job data.