NEWEarly access for 2027 hiring is open →

Every decision, on the record

RecrewOS quotes. You decide.

The exact resume lines behind every screening call, the math in plain view, and a person making every decision. You can always see why.

In private development. Design partners get early access.

Join early access See how it works

Illustrative example with fictional applicants.

Why RecrewOS is different

Other tools give you a score. RecrewOS gives you the reasons.

01No black-box scores

Plain math adds up the evidence. Every number has a reason you can read.

02Nothing made up

Every quote is checked against the resume. If it is not in the resume, it is not in the score.

03Unsure? A person decides.

Missing or conflicting evidence goes to a recruiter. Never a silent rejection.

04One record for recruiters and HR

The call, the proof and the explanation live in one place HR can open later.

The bridge between recruiting and HR

Recruiters move fast. HR gets the full story.

Read the proof, not the pile: a few quoted lines for each skill instead of pages of resume. And the same record answers HR when someone asks why.

For recruiters
  • Quoted proof for each skill, not whole resumes to reread
  • Clear cases sorted; hard ones flagged for a person
  • Gaps and mismatched dates surfaced before the interview
  • Names and contact details hidden while the evidence is read
For HR and People leads
  • Every decision with its evidence, reviewer and reason
  • The same locked criteria for every candidate in a role
  • A plain explanation, ready when a candidate asks why
  • One place to look instead of emails, notes and spreadsheets

Verified mechanics in our offline synthetic workflow. No real-resume accuracy claim.

0scores written by the AI. Recorded extraction output proposes quotes; plain code does the math.
100%of quotes shown are located in the source resume, or they are dropped.
8named reasons a candidate is routed to a person, so nobody is rejected silently.
1,000synthetic records in the replay verification suite.
Localfictional data only. Customer hosting and retention controls are planned.
Follow one resume

What happens to an application, step by step.

Meet Jordan, who applies for a Data Engineer role. Here is what RecrewOS does with that resume, and why anyone can look back later and see exactly how the decision was made.

  1. The resume arrivesJordan's name, email and phone number are hidden before anything reads the resume.
  2. The proof is pulled outRecrewOS finds the exact sentences that show each skill the role asks for, and quotes them word for word.
  3. A gap goes to a personOne required skill is never mentioned. Instead of guessing or rejecting, RecrewOS asks a recruiter to look.
  4. A person decidesThe recruiter reads the quotes, makes the call and writes down why.
  5. The why is savedA plain explanation is drafted from what was recorded, so the reason is never lost.

Jordan is fictional. This is how the workflow works today on sample data; sending explanations to candidates is still being built.

The problem

Someone asks why they were rejected. Can you answer?

Today the answer is a match score, a hurried note and a rubric nobody saved. RecrewOS keeps the proof, the math and the person who decided, together.

ExampleA scattered "why" request
  1. An 82% match scoreNo reasons attached. Nobody can say which words in the resume moved it.
  2. A recruiter's note"Not senior enough." Written in a hurry, in a different tool.
  3. A rubric that changed last monthNo version history, so the criteria used on the day are lost.
  4. A rejection emailGeneric, sent automatically, approved by nobody in particular.
  5. Three weeks of diggingWhen legal or the candidate asks, someone reconstructs it from memory.
With RecrewOSThe offline workflow
  1. Quoted evidenceThe exact resume lines behind each criterion, with location, split into supporting, contradicting and missing.
  2. Arithmetic you can redoLevel times weight for each criterion, summed in the open.
  3. The rubric version in forceLocked when it was approved, with the reason each criterion is job related.
  4. A named reviewerWho decided, when, and what they saw.
  5. Inspectable historyReview the recorded evidence and decisions locally. Portable exports and cryptographic verification are planned.
Product tour

From job description to reviewable decision.

A preview of the workflow we are building. Evidence location, arithmetic and human routing run offline; the planned import and rubric-drafting screens are illustrations.

Illustrative example. Fictional people, roles and records; not live product screens.

Approve the rubric before anyone is scored.

Drafting a rubric from a pasted job description is planned. The offline workflow already uses versioned rubrics. The proposed drafting screen will let a reviewer define criteria, weights, requirements and job-related reasons before approval.

  • In development: the internal offline check blocks "recent graduate" and warns on "culture fit". It is not a fairness guarantee.
  • Required criteria with no evidence trigger human review; they never silently determine rejection
  • Every change creates a new version; old decisions keep the version they used
Data Engineer · rubric draft v4
CriterionWeightRequiredWhy it is job related
Data pipelines3YesOwns daily ingestion into the warehouse
Production engineering3YesOn call for services the team ships
Evaluation practice2YesMeasures data quality before release
Ownership signals1NoLeads projects with unclear scope
Blocked: "Recent graduate" reads as a proxy for age. Remove it or rewrite it as a skill.
Check: "Culture fit" is vague and often hides bias. Describe the behavior you need.
Save draftApprove v4
Evidence

Every claim points to a line.

Match scores ask you to trust a number. RecrewOS shows you the sentence. Each quote is checked against the resume before it is shown, and any quote that cannot be found is dropped and counted against confidence.

  • Supporting, contradicting and missing evidence, kept apart
  • Character-level location for every quote
  • Evidence corrections retain the original in the offline record
“Led the migration of 14 batch jobs to streaming with no downtime.”SUPPORTS · PRODUCTION ENGINEERING
Arithmetic

A score you can redo by hand.

There is no gauge, badge or grade anywhere in RecrewOS. The total is level times weight for each criterion, added up in plain view, so a hiring manager can disagree with a specific line instead of a mystery number.

  • Weights are set by you, in the rubric, not by the model
  • Versioned rubric weights make score changes inspectable offline
  • Same inputs always produce the same score
pipelines 3 × 39
production 2 × 36
evaluation 0 × 20
ownership 2 × 12
total17 of 27
Human review

Uncertainty goes to a person.

When evidence is missing, contradictory or unreadable, RecrewOS does not guess. It opens a task with a named reason and puts it in front of a reviewer. A failed requirement is visible to the reviewer; "unknown" routes to a human. People make the hiring decision.

  • Eight specific routing reasons, each with its own task
  • Re-extraction creates a new result and retains the original. Closing a task does not change evidence or scores.
  • Candidate delivery is planned; no messages are sent by the current workflow
missing_evidenceEvaluation practice
contradictionTenure vs title
parse_failurescan_0091.pdf
possible_duplicateCandidate 0421
Every role

Built for every role you hire.

Each role gets its own rubric: the skills that matter, how much each counts and why it is job related. It is locked before anyone is scored, so every candidate for that role is judged on the same rules.

  • Separate criteria and weights for each role
  • A change creates a new version; past decisions keep theirs
  • PlannedSetting up your own roles from a job description
Data Engineerrubric v4 locked
Product Designerrubric v2 locked
Account Executiverubric v1 locked
Every resume

Any resume, same fair rules.

PDF, Word or spreadsheet, every candidate is read against the same criteria. Files that cannot be read are flagged for a person instead of quietly dropped.

  • PDF, Word and CSV, tested on sample files
  • Scanned, locked or broken files become named tasks
  • PlannedAn upload screen for your own resumes
PDFpriya_shah.pdfread
DOCXjordan_lee.docxread
CSVats_export.csv38 rows read
PDFscan_0091.pdfneeds a person
Security and privacy

Candidate data, handled like evidence.

These are planned safeguards, not available customer features. Today the application runs locally on fictional data. Security, hosting and data-handling controls must be completed before any customer-data pilot.

Customer isolation PlannedSeparate customer environments, databases and backups are a proposed hosting design. They are not provisioned today.
Encryption PlannedPer-candidate encryption and managed keys need implementation and security review before customer use.
Erasure controls PlannedKey-destroying erasure and backup handling are design goals. Retention periods and deletion guarantees are not established.
Provider review PlannedNo live AI provider is connected. Before transferring customer data, we will review provider terms, training restrictions and data-processing agreements.
Audit signing PlannedThe local event history is inspectable. Hash chaining and signed checkpoints remain planned security work.
Access and export PlannedCustomer accounts, multi-factor sign-in and PDF or JSON exports are not available yet.
Who it's for

One record, four kinds of reader.

HR and People leads

See the reasoning behind every screening decision in one place, instead of piecing it together from emails and notes.

Recruiters and talent teams

Read the quoted proof instead of whole resumes, and hand the unclear cases to a person.

Legal and compliance

A step-by-step history of who decided what, and why, to walk through with counsel. Signed exports are planned.

Hiring managers

Know why a candidate reached you and which gaps to ask about in the interview.

Built for the rules

New AI hiring rules. Plan for the record.

Rules for AI in hiring are spreading across states and countries. Most point at the same things: tell candidates, explain decisions, keep a person in the loop and keep the records. We are designing RecrewOS around those. Your counsel decides which rules apply to you.

Notice to candidatesPlannedrecord which notice version a candidate received and when.Decision record
Explaining a rejectionIn development: internal offline explanation drafts from record fields. Delivery is planned.Explanation approval
Human reviewThe offline workflow records human decisions. Customer communication is not connected.Review queue
Record keepingPlannedretention policies, PDF and JSON exports and erasure controls. No three-year retention promise today.Export and erasure

RecrewOS is in development; it does not make you compliant and is not legal advice. Responsibility stays with the employer.

Known limitations

We open with what it gets wrong.

Synthetic verification checks engineering behavior, not hiring accuracy or fairness. No real-resume accuracy study has been completed, and the customer workflow is still in development.

ON THE RECORDKL-01

In synthetic-file tests, scans without a text layer are not read. They become named tasks for a person; no customer upload screen exists yet.

ON THE RECORDKL-02

The proxy-term check is in development as an internal offline feature. It is a first filter, not a fairness guarantee; rubric review still needs a person.

ON THE RECORDKL-03

Real-resume accuracy is unmeasured. Before making accuracy claims, we will evaluate against practitioner labels and report sample sizes, uncertainty and limitations.

ON THE RECORDKL-04

RecrewOS records the screening it performs. It does not yet capture other tools' decisions.

Design partners

Help shape it. Review the workflow.

We are inviting employers to help shape RecrewOS around their screening workflows. In private development. Design partners get early access; customer-data pilots depend on security, legal and product readiness.

Join early access
What you get
  • Plannedcollaborative role and rubric setup
  • Planneda role-specific evaluation after data-use approval
  • A say in what we build next
  • Proposed pilot pricing, with terms to be agreed
What we ask
  • A discussion of two or three roles; no real resumes needed now
  • A 30-minute check-in every two weeks
  • Honest feedback, including what does not work
“An unexplained score adds a step instead of removing one. The explanation is the product.”
THE RECREWOS DESIGN PRINCIPLE
Pricing

Start with a pilot. Keep the record.

Proposed pricing for future pilots, not a live offer. Availability, limits and terms are not final. Customer isolation, encryption, exports and integrations are planned, not included features available today.

Pilot planned

$2,500
one time, 90 days, credited to year one
  • Proposedup to 3 open roles
  • Proposedup to 1,000 candidates
  • Plannedcollaborative rubric setup
  • Plannedrole-specific evaluation
Join early access

Team planned

$350
per reviewer per month, billed annually
  • Role and candidate limits to be agreed
  • Customer decision workflow in development
  • Plannedexport, retention and erasure
  • Plannedsigned daily checkpoints
Join early access

Organization planned

Custom
for multiple teams or regions
  • Proposedmultiple-team support
  • PlannedATS connection
  • Plannedsecurity review support
  • Plannedcustom retention policies
Join early access
Questions

You've got questions. We've got answers.

Does the AI decide who gets rejected?

No. In the offline workflow, recorded extraction output proposes quotes. Each is located in the source or dropped, and plain code computes scores from rubric levels and weights. Live model extraction is planned. People make hiring decisions. Customer messaging is not connected.

Will this make us compliant with AI hiring laws?

No. RecrewOS is in development and is not legal advice. Notice, customer delivery and retention controls are planned. Your counsel must determine applicable obligations and whether the completed workflow supports them.

How is this different from the AI matching in our ATS?

Our focus is inspectable evidence, visible arithmetic and a human decision record. The current workflow is an offline demonstration, not a connected ATS service. ATS integrations are planned.

Where is candidate data stored?

Today the application runs locally with fictional data only. Customer hosting, isolated instances, per-candidate encryption and key-destroying erasure are planned and need implementation and review before customer data is accepted.

Does the AI model train on our data?

No live provider is connected and no real candidate data is being sent. Before any customer-data transfer, we will review provider terms, training restrictions and data-processing agreements. There is no contracted provider or customer DPA today.

What happens with resumes it cannot read?

In synthetic-file tests, unreadable, password-protected or corrupt files become named review tasks rather than silent rejections. Possible duplicates also route for review. A customer upload screen is not built yet.

Do you integrate with our ATS?

Not yet. ATS connections and a customer upload screen are planned. Parsing and import checks currently run offline with synthetic files.

How accurate is it?

We have not evaluated accuracy on real resumes. Synthetic verification, including a 1,000-record replay, checks pipeline behavior, not whether it selects the best hire. Practitioner-labeled evaluation is needed before making accuracy claims.

How long does setup take?

Customer setup time has not been measured. Job-description drafting and customer uploads are planned. The current demonstration uses fictional records; pilot scope and timing will be agreed after readiness checks.

RecrewOS

Quoted evidence, visible arithmetic and human review. In private development; audit signing is planned.

Product
Product tourEvidence and arithmeticSecurityPricing
For teams
HR and PeopleTalent teamsLegal and complianceDesign partners
Trust
Built for the rulesKnown limitationsFAQJoin early accesshi@recrewos.com
RecrewOS is in development and does not provide legal advice.© 2026 RecrewOS