For recruitment agencies and tech companies

The hiring engine
that shows its work.

Source, match, submit, interview, offer and place — every stage in one pipeline, every recommendation backed by evidence you can check. Run it in ApplyForge, or plug it into your own product with our API.

Stage 2 of 6 · Match

A ranked shortlist that explains itself.

Paste a brief and every candidate is scored 0–100 against it, with the reasons, the gaps, and what to check on the call.

  • Explainable fit score per candidate
  • Gaps flagged, never assumed
  • Screening questions built from the gaps

API: GET /jobs/{id}/shortlist

James R. · fit 92

Senior Python Engineer · Halden Logistics

Top match
Required skills31 / 35
Profile similarity27 / 30
Experience15 / 15
GapTerraform not evidenced

Evidence · Python + FastAPI, 6 yrs — CV, roles 1–2. AWS in production — CV, role 1.

Illustrative example — select a stage above

One engine, three ways in

Built for whoever owns the hire.

Recruitment agencies

Run every desk from brief to placement.

Win more placements from the network you’ve already built, and give clients a page they actually want to open.

  • Submissions with spec-sheets and client pages
  • Offers, fees and placements tracked
  • Next-hire advisory for every client

Tech companies

Hire the strongest candidate, not the first.

Rank every applicant against the spec, brief every interviewer, and give hiring managers the reasons behind each recommendation.

  • Every applicant ranked with evidence
  • Interview briefs from prior feedback
  • Offer history in one place

Developers & platforms

Ship explainable matching in your product.

Careers sites, job boards, client portals and HR tools: send jobs and applicants in, get ranked shortlists and webhooks out.

  • REST API with scoped keys
  • Signed webhooks for every stage
  • XML job feed for boards

The full recruitment pipeline

The hard part of hiring starts after the shortlist.

Most AI recruiting tools stop at “here are some matches.” ApplyForge carries the evidence through submissions, interviews, offers and placements — so nothing is lost between stages.

Submissions

Submit with a case, not just a CV.

A readiness check flags what’s missing before anything goes out. The client-safe write-up is drafted from the CV and your screening notes — and never claims what the evidence doesn’t support.

  • Branded or anonymised spec-sheet, PDF or Word
  • Screening notes frozen into the submission
  • Client feedback per candidate, back in the role

Submission readiness · James R.

4 of 5

CV parsed and verified

Screening call notes added

Right-to-represent consent recorded

Salary expectation captured

! Notice period not confirmed

“James has six years of production Python, most recently leading FastAPI services on AWS at a payments firm [CV]. He’s open to hybrid in London [screening].”

Interviews

Every interviewer walks in prepared.

Propose and confirm slots, then generate a brief from the role, the CV and every earlier round of feedback — so stage two picks up exactly where stage one left off.

  • Proposed → confirmed scheduling
  • Brief built from gaps and prior feedback
  • Per-candidate, per-role SWOT

Interview brief · 2nd stage

Confirmed · Thu 14:00

Probe

Terraform isn’t evidenced — ask how infrastructure changes reach production.

Stage 1 panel noted thin system-design depth — go one level deeper.

SWOT

Strength · AWS in productionWeakness · no IaC shownOpportunity · platform leadThreat · salary above band

Offers

Negotiate with the whole history in view.

Every offer, counter and revision on one timeline, with talking points grounded in what the candidate actually asked for — and market data for when the budget is the problem.

  • Offer and counter-offer history
  • AI talking points for the next call
  • Placement fee worked out automatically

Offer timeline · Senior Python Engineer

Mon · Initial offer£82,000
Tue · Candidate counter£92,000
Wed · Revised offer£88,000 · accepted
Placement fee at 20%£17,600

Placements

Every placement starts the next one.

When a role closes, the runners-up are re-matched to your other open roles, and the client’s team shape suggests what they’ll hire next — with the supply you already have behind it.

  • Runners-up matched to open roles
  • Next-hire advisory per client
  • Daily brief of what needs you today

Halden Logistics · next likely hire

Placed ✓

Their team now has 4 backend engineers and no platform or DevOps role.

Platform / DevOps Engineer6 in your pool
Engineering Manager2 in your pool

Runner-up Sarah M. also fits 2 of your open roles.

All names, figures and fees above are illustrative

Partner API v1

Put explainable matching inside your own product.

Publish jobs from your careers site, post applicants in, and read back a ranked shortlist with the reasons and gaps behind every score. Stage changes reach you by signed webhook the moment they happen.

RequestResponse
GET /partner/v1/jobs/job_8kq2/applicants
Authorization: Bearer af_live_••••••••

{
  "applicants": [{
    "name": "James R.",
    "fit_score": 92,
    "reasons": ["Python + FastAPI, 6 yrs", "AWS in production"],
    "gaps": ["Terraform not evidenced"],
    "stage": "screening"
  }]
}

Jobs

Create, update and close jobs — or upsert by your own reference.

PUT /jobs/by-external-ref/{ref}

Parse a job spec

Turn a pasted brief into structured requirements.

POST /jobs/parse

Applications

Post applicants in, move stages, add notes and documents.

POST /jobs/{id}/applications

Ranked shortlists

Scores with reasons and gaps for every applicant.

GET /jobs/{id}/applicants

Public job feed

Your open roles as JSON for your site or XML for boards.

GET /feeds/{agency}/jobs.xml

Privacy built in

Candidate erasure, usage metering, rate limits and an audit log.

DELETE /candidates/{id}

Signed webhooks

job.createdjob.updatedjob.closedapplication.createdapplication.processedapplication.stage_changedcandidate.deletedcandidate.erased

Chrome extension

The engine, on every LinkedIn profile.

Open a profile, pick the role you’re hiring for, and see its fit score with reasons and gaps — the same matcher as the rest of the pipeline. One click drops them into Sourced.

  • No separate login — uses your ApplyForge session
  • Scoring is a preview; nothing saved until you click Add
  • Saved twice? The record updates, never duplicates
ApplyForge RecruiterSigned in

Lead Data Engineer — Northwind Bank

Priya S. · Strong fit

Python, Spark · Payments · gap: Snowflake

Illustrative — profile and role are fictional

AI that shows its work. People who make the call.

ApplyForge isn’t an autonomous agent hiring on your behalf. It ranks, explains, drafts and reminds — and a recruiter or hiring manager decides every step.

Evidence on every score

Reasons and gaps trace back to the CV or profile.

Never invents facts

Drafted write-ups cite their sources; unsupported claims are flagged.

Humans decide

Recruiters and hiring managers own every stage change.

Isolated workspaces

Every record is scoped to your organisation.

UK GDPR aligned

Consent, retention and erasure — in the app and by API.

No shared training

Your CVs and notes never train shared models.

Get started

Prove it on a real hire.

Agencies & hiring teams

14-Day Placement Sprint

£0

Bring one live vacancy and the candidates you already have. Run it from sourcing to offer with us beside you.

✓ Full pipeline · ✓ Chrome extension · ✓ Client pages

Start your free sprint

Developers & platforms

Partner API

Contact us

Scoped keys, signed webhooks, rate limits and an audit log. Tell us what you’re building and we’ll set you up.

✓ Jobs & applicants · ✓ Ranked shortlists · ✓ Webhooks

Talk to us about the API

No credit card · No ATS migration · UK GDPR aligned

Questions buyers ask.

Is this for agencies or for in-house teams?
Both. Agencies use client pages, spec-sheets and placement fees; in-house teams use the same pipeline with hiring managers in place of clients. The matching and evidence are identical.
Does ApplyForge decide who gets hired?
No. It ranks, explains and drafts. Every score shows its components, every gap is flagged rather than assumed, and a person makes every decision.
Can we use it without leaving our ATS or careers site?
Yes. The Partner API lets your careers site publish jobs, post applicants in and read back ranked shortlists, with signed webhooks for every stage change. There is also an XML job feed for boards.
How does the Chrome extension sign in?
It has no login form — it uses your ApplyForge session from the web app. Scoring a profile stores nothing until a recruiter clicks Add.
Is our candidate data shared or used to train models?
No. Each workspace is isolated, data is never visible to another customer, and your CVs and notes are not used to train shared models. Candidate erasure is available in the app and by API.
What does it cost?
The 14-day placement sprint is free with no card. For the Partner API, contact us. See pricing for plans after the sprint.

From first look to signed offer.
With the evidence at every step.