AI Hiring SaaS

01 Context

B2B hiring SaaS at Zeero.us, during a 2025 client-engagement period. Contribution sat at the intersection of recruiter workflow, ranking quality and analytics.

02 Problem

High application volume made screening slow and inconsistent. Recruiters needed faster shortlisting without a drop in ranking quality.

03 Users / Stakeholders

• Recruiters and hiring teams
• Hiring managers waiting on qualified pipelines
• Engineering and data counterparts owning ranking, workflow and analytics

04 My Scope

AI hiring engine: resume screening, job–candidate matching, recruiter workflow, ML ranking and recruitment analytics.

05 Product Decisions

• Centre the product on ranking quality and recruiter workflow, not a generic chatbot over resumes
• Expose match scores and analytics so recruiters could inspect rather than blindly accept model output
• Learn from recruiter feedback so ranking improved inside the workflow rather than as a sidecar model

06 Technical Constraints

Ranking quality depended on ML candidate ranking and on fitting the workflow recruiters already used. Integration with existing HR systems constrained how much of the process could be redesigned at once.

07 Trade-offs

• Automation versus recruiter control
• Screening speed versus ranking precision
• Model complexity versus inspectable scores in the workflow

08 Measurement

Ranking accuracy and candidate screening time.

09 Outcome

+18% ranking accuracy and −35% candidate screening time.

10 Reflection

The useful product question was not whether AI could read resumes. It was whether recruiters could move faster without losing ranking quality, and whether that quality was visible in the workflow.