Lighthouse Projects · HR
Recruitment automation: AI-assisted candidate screening
Less admin for recruiters, with the decision always in a person's hands.
Time to hire
14 d
Before (manual)
38 d
Problem
Recruiters spending time on manual CV triage, contacts and scheduling. Qualified candidates lost to poor follow-up. Inconsistent candidate experience.
Solution
Pipeline with AI-assisted pre-screening that the recruiter reviews before deciding, communication agent for personalized messages and scheduling, ATS and calendar integration. Recruiter funnel dashboards.
KPIs
- Time from open role to hire
- Candidate response rate
- Recruiter time spent on admin
- Candidate experience
How it is measured
Every KPI with a baseline and a target.
The numbers for this project come from your data. We measure the baseline during framing, agree the target with you before building and, in production, measure the same KPIs the same way.
- 01BaselineMeasured during framing, with the data of the process as it runs today.
- 02TargetAgreed with the case owner before building, starting from the baseline.
- 03In productionThe same KPIs, measured the same way, compared with the baseline.
- Time from open role to hire
- From opening the role to the offer being accepted, from the ATS dates.
- Candidate response rate
- Candidates who reply to each contact, divided by those contacted, per stage.
- Recruiter time spent on admin
- Hours per week on screening, contacts and scheduling, measured by sampling before and after.
- Candidate experience
- A short survey at the end of the process, for hired and not hired candidates.
Before starting
Prerequisites and risks.
Prerequisites
- An ATS with an API or export, and the recruiters' calendars.
- Screening criteria written and approved by HR before building, tied to the role's requirements.
- The data protection impact assessment (GDPR, Article 35) done before the pilot.
- Candidates informed that screening uses AI (AI Act, Article 26(11)).
Risks and how we handle them
High risk under the AI Act
AI that analyses and filters applications or evaluates candidates is high-risk (AI Act, Annex III, point 4(a)). Under Regulation (EU) 2026/1744, those obligations apply from 2 December 2027, and the pilot is designed with them from the start.
Human oversight
The AI ranks and summarises applications. Whoever advances or rejects a candidate is a recruiter, who sees the criteria used and can override the suggestion. Oversight sits with people who have the training and authority to decide (AI Act, Articles 14 and 26(2)).
Automated decisions (GDPR, Article 22)
No candidate is rejected by an automated decision alone. The GDPR gives candidates the right not to be subject to a decision based solely on automated processing that significantly affects them, and Recital 71 gives e-recruiting without any human intervention as an example.
Bias in the criteria
Criteria learned from past hires can repeat their biases. The criteria are written and reviewed by HR, and in the pilot the screening is compared with the recruiters' decisions.
Other projects · HR
Want to see this in your operation?
90 minutes with you and your team. We leave with 3 concrete opportunities and estimated payback.