ZekeStack

Recruitment operations

AI Candidate Intake and Shortlisting System

Automates candidate intake and first-pass review for businesses handling frequent hiring, high applicant volumes, and repeated screening workflows.

Business outcome

Reduce manual applicant screening and speed up hiring response times while keeping human approval and auditability in the process.

Use case fit

Best fit for businesses that handle frequent hiring, high applicant volumes, or repeated candidate screening across multiple roles. Useful for recruitment agencies, care providers, hospitality groups, logistics teams, retail operators, and other organisations where hiring workflows create significant operational admin.

Operational workflow

The workflow diagram shows how the automation moves work from intake to review, downstream updates, and auditability.

Candidate intake
AI evaluation
Recruiter review
Shortlist decision
ATS/CRM update
Audit log

Reference architecture

The architecture view uses neutral system blocks to show data flow, integration boundaries, review points, and operational logging.

Sources

SRC

Candidate sources

Email, form, job board, LinkedIn export

Automation core

UI

Intake UI

Candidate profile capture and role selection

AI

AI scoring layer

Fit score, summary, matched skills, gaps

REV

Recruiter dashboard

Shortlist, reject, or request more information

Integrations

API

ATS/CRM connector

Candidate status, notes, pipeline updates

MSG

Notification layer

Recruiter alerts through email, Slack, or Teams

Governance

LOG

Audit log

Traceable automation and human-review events

Technical stack

  • Next.js
  • TypeScript
  • Tailwind CSS
  • Mocked AI scoring
  • OpenAI-ready API layer
  • ATS/CRM integration path

Integration path

  • CV parsing
  • OpenAI evaluation route
  • Postgres or Supabase storage
  • Bullhorn, Greenhouse, Lever, HubSpot, or Zoho Recruit
  • Email, Slack, or Teams notifications

Implementation notes

  • The public demo uses mocked AI to keep walkthroughs stable.
  • A production build can replace the mock layer with a real model and role-specific scoring rules.
  • The architecture keeps recruiters responsible for final decisions.