Search jobs with a resume. Get dramatically better matches.
Upload a resume, get ranked matches from millions of active listings indexed direct from 200,000+ company career pages. The resume carries the signal. Our data and infrastructure turn it into the highest-quality job match available through an API.
The Unlocks
A resume is the highest-signal query you have.
When a user has a resume, the match upgrades. Years of work history, specific projects, technologies, seniority, domain. All the signal a keyword query throws away. We turn that resume into a 768-dimensional vector and rank it against millions of jobs from the cleanest corpus on the market. The resume is the unlock. The infrastructure is what makes it work.
Resumes carry meaning. Use it.
Years of work history, specific projects, seniority, domain expertise. A resume encodes all of it. Our embedding model captures that signal and ranks it against millions of active listings. Better signal in. Better matches out.
The best signal needs the best corpus.
A great match against bad data is still a bad match. We index direct from 200,000 company career pages, never third-party aggregators. No duplicates. No agency reposts. No expired listings lingering for weeks. The cleanest job dataset, refreshed every few hours.
Three API calls. Not three months.
Embedding model. Vector index. Freshness pipelines. ATS coverage across 20 platforms. Location normalization. Salary parsing. We built the plumbing. You ship the feature this week.
Two calls. Ranked matches.
Vectorize the resume.
POST a resume file to /v2/resumes/embed. We parse the structured fields and generate a 768-dimensional vector via socrates v2, our proprietary embedding model trained on resume-to-job-description pairs. Returns an artifact_id plus the parsed resume. Accepts PDF, DOC, DOCX, TXT, HTML up to 5MB.
Read the DocsRun neural search.
POST to /v2/jobs/neural-search with vector.artifact_id from step 1. Combine semantic ranking with structured lexical filters in the same call: titles, keywords, geo_locations, salary, experience, visa, industry, days_ago, sort_by. Returns ranked job objects with vector_score from 0.0 to 1.0.
Read the DocsOr use vector search.
For pure semantic similarity without lexical filtering, hit /v2/jobs/vsearch with search_type: 'resume' and the same artifact_id. Configurable accuracy (low / medium / high) and top_k up to 500. Best for raw similarity ranking before downstream re-ranking.
Read the DocsTry it on your own resume.
Upload a resume. We will embed it and run the match against millions of active listings. PDF, DOC, DOCX, TXT, or HTML up to 5MB.
Upload your Resume
Upload a resume to see ranked matches from the live index. No signup required.
Three calls, any
language.
// Response payload
{
"company": {
"name": "Stripe",
"headcount_growth": "+12%",
"open_roles": 142
},
"jobs": [
{
"title": "Staff Engineer",
"stack": ["Ruby", "React"],
"salary_range": {
"min": 220000,
"max": 350000
}
}
]
}The match is only as good as everything underneath it.
Other APIs
Hirebase
Built for products that depend
on getting the match right.
Best matches for your resume.
Show users semantic matches without building the embedding stack. Replace generic search with semantic ranking and watch session depth go up.
Surface the next role.
Recommend relevant external openings to candidates already in your pipeline. Increase placement rates without sending users elsewhere.
Power "what to apply to next."
Match resumes to roles your users would never find with keyword search. Build the recommendation product career platforms have promised for a decade.
Match candidates to live reqs.
Match candidate profiles to active requirements at scale. Find the senior engineer whose resume says ML pipelines for the role asking for data infrastructure.
A matching API is only as good as the corpus.
Most resume matching products run on top of aggregated job data. They inherit every problem of the upstream feed: duplicates, stale listings, recruiter reposts, missing fields. We bypass the entire aggregator layer.
We index directly from 200,000+ company career pages and 20+ ATS platforms including Greenhouse, Lever, Workday, Ashby, iCIMS, SmartRecruiters, BambooHR, Taleo, Dayforce, and Workable. Every listing is keyed to a single canonical company entity. Every job is re-verified against its source multiple times per day. Every description is enriched by LLM into 50+ structured fields before it enters the index.
The match is the last step. The data underneath is what makes it work.
Specs
Capabilities
Pricing that scales with you.
Free
For production integrations. Up to 100 results per page across Search, Neural Search, and Resume endpoints.
Growth
For production integrations. Up to 100 results per page across Search, Neural Search, and Resume endpoints.
API Starter
For production integrations. Up to 100 results per page across Search, Neural Search, and Resume endpoints.
Enterprise
For production integrations. Up to 100 results per page across Search, Neural Search, and Resume endpoints.
Questions, Answered.
Build on the layer everyone else is trying to reach.
The embedding model. The vector index. Millions of jobs from 200,000+ companies, re-verified hourly. Three API calls away.
