Here are my slides & prompts from the always fun NJA meetup.
Download the slides (PDF, 2.5 MB)
This was a lightning talk to help folks get an LLM working with candidate data, and making sure the results are as hallucinations free as possible.
We’ve been running hard with AI this year. We built on our own ATS/CRM (called Lloyd), have a developer on the team and believe that LLMs will be a force for good in hiring. We wrote code to bring our data sources together - our ATS/CRM, emails, LinkedIn, Slack channels where we work with clients. Only recently have we started to really see the benefits of having everything in the one place.
AI is an infinite workforce for very specific things and it’s our job to find those things. With 14 years of history in place, one of those specific things is sharing some incredible context on our clients. For new clients, a human + LLM speed runs the getting to know you process.
Another specific thing - hiring rubrics. Building and working towards a hiring rubric happened for retained exec roles, but IC positions moved too fast. Now, every role gets a rubric. Even better, it’s updated automatically as our clients interview candidates and share feedback (often in Slack). When a rubric changes, Lloyd re-reviews applicants in case someone was screened out before who could be included now. It does a fresh database search and checks external sourcing tools too. Even a well staffed recruitment team would not have done that sort of thing.
I can talk about this stuff all day. If you’re building teams and think we can help - please get in touch.
