The requirements that decide who gets hired are rarely on the job description. “Overachiever.” “High agency.” “Scrappy.” Hand those to an AI and it matches titles and keywords, then fills your pipeline with candidates your team rejects.

Your best recruiter learned what those words meant by sitting in intakes and debriefs. Your AI got a job description. Lorenzo Casanova will show how to onboard your AI the same way you’d onboard a star recruiter: define every attribute, exception and unwritten rule once. The difference is the AI holds all of them at the same time, which no recruiter can.

He’ll also take it past the profile: validating licenses, reviewing GitHub work, checking whether someone’s publications matter, evaluating portfolios.

No specific tool required. Bring whatever AI recruiting system you use today.

Key Takeaways

How to break an undefined attribute into evidence a profile can show.
When a signal misfires (a title, a keyword, a side project), refine what counts before you drop it.
Hard requirements vs. preferences, why missing evidence means unknown, and how to stop your AI from overweighting one detail.
Test before you trust: compare its calls to hires you already made, run it twice, and read the reasoning, not just the score.