HR Application Screening

Problem

Shortlisting job applications against a role specification is time‑consuming and subjective, particularly where evidence is presented in free‑text formats such as CVs and supporting statements.

Delivery

The Innovation team developed an experimental system that extracts requirements from job descriptions and related documentation to generate a scoring rubric, against which applications can be assessed in a similar way to rubric‑based essay marking.

Outcome

The work demonstrated that applications can be scored consistently against role‑specific criteria, while highlighting the importance of carefully defined rubrics and wording, and providing a foundation for ongoing investigation into AI‑assisted screening approaches.