Scores travel with evidence, risks, and interview questions so HR can verify faster.
PressHR Platform
PressHR: evidence-led AI recruiting with human approval.
PressHR is a lightweight ATS for HR teams that want to read CVs faster without giving final decisions to automation. It brings CVs, job descriptions, scorecards, AI review, approval flow, email outbox, and audit into one controlled workspace.
CVMarkdown parse
AIScorecard review
ApprovalHR, manager, audit
Lightweight AI ATS
AI reads faster. People decide.
PressHR is built for recruiting teams that need clear candidate context, evidence, risks, and approval status without chasing scattered files, chat, and email. AI assists analysis; HR still reviews, overrides, and submits for approval when the evidence is ready.
AI advises only; hire, reject, approval submission, or email actions still need a responsible user.
The timeline keeps reviews, overrides, approval actions, and email status traceable.
HR
CV
JD
AI
AUD
CV Parse
AI Review
Human Approval
Audit
Recruitment workflow
From CV to approval in one closed loop.
PressHR connects upload/import, parsing, AI scorecards, HR override, approval, and email outbox in one path. HR sees the real candidate state instead of guessing where the file sits or who owns the next decision.
Recruitment control flow
CV in. Evidence out. Decisions approved.
Receive CVs by job, file, or folder, then place them in private storage.
Workers build clean content and lock AI when data is not reliable enough.
AI reads CV, JD, and snapshotted criteria to produce evidence-led review.
HR checks scores, evidence, risks, and can override with a reason.
Managers approve, request changes, or forward to the next level.
Candidate email goes through outbox instead of extending business requests.
Evidence-led scorecard
Scores must explain why.
PressHR does not reduce recruiting review to a cold number. Each review needs dimension scores, verifiable evidence, risks to clarify, and interview questions tied to criteria. Scorecards are snapshotted per job so old reviews do not drift when new criteria change.
Quickly see fit across each criteria group.
Tie scores to points HR can verify in the CV.
Surface missing data or points to clarify in interviews.
Let responsible users adjust recommendations with reasons.
Multi-level approval
Sensitive decisions need a clear timeline.
PressHR approval lets HR submit a candidate to the responsible person, managers approve or request changes, and the next level can be added when needed. AI draft only helps prepare forwarding notes; it does not overwrite the original review or close the candidate.
AI draft
Approval comment draft
Draft forwarding notes so approvers read faster, while real users still confirm.
HR submits after evidence review
Manager approves / requests changes / forwards
Final hire / reject creates outbox and audit
Governance
Trust starts with honest security language.
CVs and job descriptions are sensitive data. PressHR keeps original CVs in private storage, limits actions through RBAC, protects mutations with CSRF, records audit logs, and sends email through outbox. If a customer enables a cloud AI provider, data may leave the system for processing under approved configuration; only local/private AI deployments fit a no-provider-egress claim.
No public object URLs for original CVs.
Sensitive mutations require permissions and valid tokens.
Reviews, overrides, approvals, and email stay traceable.
Cloud AI is described as customer-approved egress.
PressHR deployment
Start with the core hiring flow, then expand when the workflow is clear.
Go + HTMX keeps screens lean, PostgreSQL holds data and queues, and workers handle parse/AI/email outside request paths. PressUp can scope deployment by department count, CV types, approval levels, email templates, and AI provider requirements.