TA Crew: Candidate Screening
TARA REYES
Every applicant scored, every call still the recruiter's.
A large APAC employer, such as a bank, mobile operator or insurer, receives about 10,000 applications a month across hundreds of open requisitions, each with its own hard gates and criteria. Recruiters cannot apply them consistently at that volume, and candidates who miss one role but fit another are rarely redirected. Tara scores every applicant against the requisition's own criteria, while every advance, redirect or decline stays a recruiter's decision with the evidence on record.
- 0 hrs
- Freed a month, worth ~$22.5K
- Next day
- To screening outcome, not 3–5 days
- 0%
- Near-misses checked for other roles
Capabilities
Four steps, every morning.
Lists who is waiting
Runs daily at 06:00 and lists every candidate pending screening, with the requisition and the level applied for.
Scores against the requisition
Fetches that requisition's hard gates, criteria and threshold fresh, parses the resume and scores every candidate, tagging each match exact or semantic.
Looks for another fit
When a candidate does not qualify, checks other open requisitions at the same level or below and records a suggested redirect or rejection.
Hands the evidence to the recruiter
Writes the score, matched skills and notes back to the candidate record and notifies the recruiter in the team's messaging tool.
Where it is deployed
INDUSTRIES
TEAMS
Governance
Tara holds no screening criteria of its own: every run reads the requisition's guidance fresh from the recruitment platform, and the recruiter decides every outcome.
- Autonomy tier
- Advise. Scores every applicant and recommends an outcome. The recruiter advances, redirects or declines, and can ask for an on-demand rescreen of any candidate.
- Human Principal
- A named person owns performance and approves scope. Every AI Coworker reports to a human.
- Off limits
- Never changes a candidate's stage. Never applies criteria of its own. Every advance, redirect or decline stays with a recruiter, with the evidence recorded for audit and fairness review.
- Audit
- Every score, matched skill and recruiter decision logged with its evidence, for audit and fairness review.
Built on 
Candidate records, resumes and evidence stay in your own AWS account under a customer-managed key. Modelled platform cost is about $0.12 per application.
Deployment
This runs as an OBZ-led engagement in your own AWS account. Tara reads each requisition's screening guidance, so a new role family is configured in your recruitment platform; the coworker itself stays the same.
Scoped to your requisition and application volume.
Talk to us about deploying Tara ReyesWhat we need from you
- Access to your recruitment platform or ATS
- Screening criteria set for each role
- A business messaging tool and an AWS landing zone
- Your named Human Principal
