Resume & job matching for job marketplaces · US & EU

The right job for every candidate. The right candidate for every job.

Counting AI connects to the resumes, job posts and activity data already in your cloud and matches them in both directions: the best jobs for each job seeker, the best candidates for each recruiter, and the reasons behind every match.

Both sides
jobs for candidates and candidates for jobs, from one understanding of your data
Explained
every match comes with its reasons, quoted from the resume and the job post
Your cloud
resumes and job posts stay where they are; results are written back to them
Built for hiring platforms
  • Job boards
  • Job aggregators
  • Staffing & recruiting firms
  • Recruiting software

Markets

Built for US and EU hiring.

Two of the most closely regulated places to use AI in hiring. We design for their rules from the start.

US

United States

For job boards, staffing firms and recruiting software serving US employers.

  • NYC Local Law 144
  • State AI hiring laws
  • EEOC guidance
  • CCPA / CPRA

EU

European Union

Run in EU regions, in the language of each market.

  • GDPR
  • EU AI Act
  • Data stays in the EU
  • Local languages

One system, two rulebooks

The same matching runs in the region where your data lives, with that market's rules built in.

How it works

Every resume and every job, read side by side.

We work on the data already in your cloud, and connect every candidate to the jobs they fit, and every job to the people who fit it.

In your cloud

Resumes & profiles
Job posts
Applications & outcomes
Search & click activity
Screening calls & video
Counting Authorized access
  1. 01Parseskills, titles, seniority, location and pay, on one taxonomy
  2. 02Retrievea shortlist from your whole pool of people and jobs
  3. 03Rankeach resume and job post read together by frontier models
  4. 04Explainthe reasons for every match, and what is missing

Written back to your cloud

Job recommendations
Job alert emails
Candidate shortlists
Match reasons
Skills & pay tables

Services

What we build for you.

Four kinds of systems, each tuned to your marketplace and your data, and run by our team.

Resume & job parsing

Every resume and job post turned into clean, structured records: skills, titles, seniority, location and pay, on one taxonomy.

For exampleA two-page PDF resume and a free-text job post become records you can search, filter and compare.

Job recommendations

Ranked jobs for every job seeker, on your site, in your app and in job alert emails, each with a one-line reason.

For example“Matches your five years of Kubernetes and the on-call experience this role asks for.”

Candidate matching

Ranked candidates for every job, and search in plain language, so recruiters reach the right people first.

For example“ICU nurses within 20 miles, open to night shifts, licensed in California.”

Job quality & insights

Duplicate, expired and suspicious posts flagged, and skills and pay trends across your marketplace.

For exampleThe same job reposted under different titles, merged into one listing.

Why Counting

Three things we never compromise on.

Scale

Billions

of tokens processed by frontier models every week, reading resumes, job posts and screening calls side by side.

Accuracy

Frontier

The right frontier model for every step, upgraded as new ones ship. Every match comes with its reasons.

Privacy & fairness

Private

Resumes stay in your cloud. We read them only with your authorization, never train models on them, and test every ranking for bias.

Works with

  • Resumes & profiles
  • Job posts
  • Applications & outcomes
  • Search & click activity
  • Screening calls & video
  • Recruiter feedback

Deployment

  • Connected to your cloud
  • Deployed in your project
  • On-premises

Security & privacy

  • Encryption with your keys
  • Private networking
  • SSO & role-based access
  • Audit logs
  • Data residency
  • Deletion on request
  • NDA before data

Fair hiring

  • People make every hiring decision
  • Explainable rankings
  • Bias testing
  • Records for bias audits
  • Consent before interview analysis

Markets

  • United States
  • European Union

Languages

  • English
  • German
  • French
  • Spanish
  • Italian
  • Dutch
  • Other EU languages

How we work

Proof first, then production.

01 1–2 weeks

Discovery

Map your resume and job data, your funnel and the metrics that matter.

02 4–6 weeks

Proof of concept

Matches on your real data, A/B-tested against your current system on clicks, applies and recruiter replies.

03 2–4 months

Build

Production in your environment, wired into your site, app and emails.

04 Ongoing

Operate

Monitoring, bias testing and model upgrades.

About

Silicon Valley engineers who love finding knowledge in data.

A resume and a job post describe the same work from two sides, written by people who have never met. We build systems that read both closely, and show why every match was made.

Team
Veteran AI and data engineers
Expertise
LLMs · machine learning · data mining · cloud
Markets
United States and European Union
Based in
Sunnyvale, California

Contact

Tell us where your matches fall short.

And the data you already have. A founder will reply.

Email
hello@counting.ai
Office
Sunnyvale, California
Built for
Job boards, job aggregators, staffing firms and recruiting software
Markets
US and EU
Job seekers
We work with companies, so we can't review individual resumes or applications.