Buzzwords, rewritten project bullets, and role-specific phrasing make weak and strong candidates harder to separate by reading alone.
Turn 100 ITresumesinto 10 proofpackets.
AI made tailored IT resumes cheap. Acruit helps staffing teams verify candidate claims, surface missing proof, and hand hiring managers a shortlist they can trust.
AI resume volume is rising faster than recruiter trust.
Every applicant can now generate a polished, keyword-matched IT resume in minutes. That does not mean the candidate is weak. It means the resume is no longer enough evidence to submit with confidence.
They do not want another AI match score. They want to know which claims are verified, which claims are missing evidence, and what risk remains.
GitHub links, portfolios, credentials, work samples, timeline conflicts, and duplicate patterns are checked inconsistently under deadline pressure.
Pay for the batch, not the software promise.
The first sale should be a concierge proof run: one real IT role, one noisy applicant export, and a concrete packet batch your team can compare against its current review process.
JD, criteria, ATS CSV, resumes, candidate URLs, and any work samples already in the funnel.
Code footprint, portfolio, credential links, duplicate patterns, missing work samples, and recruiter review flags.
A small shortlist with proof, open questions, candidate-safe follow-ups, and a decision audit.
See the applicant pile become proof packets.
The demo shows a larger market-scale batch. The paid pilot starts smaller: one IT role, up to 100 applicants, and a measurable proof packet handoff.
Acruit proof workflow animation
The animation starts with 244 applicants, narrows them to eight focus packets, shows Acruit inspecting raw evidence with a lens, and assembles three Hiring Manager Review Packets for recruiter review.
Market scale is grounded in Greenhouse’s 2026 benchmark of 244 applications per job in 2025. The 8 focus-packet review stage is consistent with Ashby’s 3.6–4.7% interview conversion range. Scenario data remains synthetic.
Only verify claims that matter.
The Musk audit says the unit is not the resume. It is the claim. The landing workflow now follows that: extract job-related claims, attach proof, ask only for missing evidence, and preserve the human decision.
Lock the role criteria
The recruiter confirms which IT claims matter: stack depth, project ownership, production experience, remote fit, authorization, and compensation range.
Split resumes into atomic claims
Acruit separates self-asserted resume text from GitHub activity, portfolio artifacts, credential links, work samples, availability, and candidate answers.
Request proof without interrogation
Candidates get a short, transparent request for the missing artifact: a work sample, project explanation, credential link, or clarification.
Send the shortlist with receipts
The hiring manager receives verified claims, evidence gaps, risk context, and a recruiter-owned next step instead of another AI rank.
The deliverable is proof, not a score.
First customers do not need a new ATS. They need fewer weak submissions, fewer re-explanations to hiring managers, and a clearer reason to ask candidates for missing evidence.
- Claim-level source links instead of one opaque authenticity score
- GitHub, portfolio, credential, and work-sample proof kept separate
- Evidence gaps become candidate-safe follow-up requests
- Every criteria edit, request, and approval remains auditable
Built to avoid the obvious traps.
The analysis pages all converge on the same constraint: trust is the product boundary. Acruit should help recruiters explain evidence, not outsource employment decisions to a black box.
Advancing, holding, or submitting a candidate always requires an attributable recruiter action. The system supports decisions; it does not make them.
Candidates can supply context, correct details, and understand why a piece of evidence was requested. The frame is proof-building, not suspicion.
Acruit reports verified, missing, and conflicting claims. It never collapses a person into a hire/no-hire score or an AI detector verdict.
Run one paid proof batch.
Send one hard IT role and a recent applicant export. Acruit returns evidence-backed packets, missing-proof requests, and a clean audit trail your recruiters can judge.