Say turnover is rising. AI can help HR pull the numbers and see where people are leaving. It takes a different kind of work to figure out whether a new manager, a brutal shift schedule, or a broken onboarding process is driving people out. That means talking to employees and making a call about what needs to change.

That’s a practical place to draw the line. Give AI the work of gathering and organizing information. Keep a person accountable for what the company does with it.

Better data can get HR to the right conversation faster

When workforce data lives in different systems, HR can spend more time assembling a report than figuring out what it says. Connecting those systems can help a team see whether departures are concentrated among new hires, on a particular shift, or in one part of the business. Benchmarks can add context about how unusual the numbers are.

That gives HR somewhere to start. If most departures are happening in the first 90 days, look at what new hires were promised and what they actually walked into. The value is in how quickly HR can get from a number on a dashboard to a conversation that could improve the way people work.

Automating a task changes somebody’s job

If an AI agent drafts the monthly workforce report, somebody still has to catch a bad comparison, check that the totals match the source data, and decide what leadership needs to see. Give that person the time and authority to do the review. Otherwise, the first draft can become the final report simply because everyone is busy.

Be specific with employees about the change, too. “This will free up your time” leaves people guessing what they’ll be expected to do with that time, whether their goals will change, and who’s responsible when the tool gets something wrong. Those are questions HR can help leaders answer before a rollout.

Put one workflow to the test

Start with something the team already does, like a recurring people analytics report. Have AI pull the numbers and draft a summary. Give HR ownership of checking the data, adding context, and deciding what to send to leadership. After a few cycles, look at the time saved after review, the corrections people had to make, and whether the report helped leadership act sooner.

Apply that same discipline to the build-or-buy decision. A prototype can show that a tool works on a sample of workforce data. Running it month after month means maintaining integrations, managing access to sensitive information, and keeping up when systems change. Count that work before deciding what’s worth building in-house.

In this episode, Joelle Emerson, co-founder and CEO of Paradigm, shares how the company moved from consulting to software and why it rebuilt its product around AI. She discusses what Paradigm’s Surface platform is designed to do with connected HR data and benchmarks, how people teams can help lead AI adoption, and where human judgment matters as the work changes.

Chapters:

00:10 Joelle Emerson and Paradigm

01:22 From employment law to building healthier workplace cultures

06:09 Why Paradigm shifted from consulting to software

09:44 Are AI agents part of the workforce?

12:35 What teams need to adapt to AI

17:00 Combining HR data with expertise and context

19:29 How AI can amplify human strengths

22:22 What employees worry about and how leaders talk about AI

24:29 AI costs, organizational design, and the role of HR

25:46 How Surface helps HR get to insights faster

29:19 When to build, when to buy, and when to hold off

32:00 Starting with one practical AI use case

35:26 How AI could reshape people teams

Today's Guest
Joelle Emerson
Co-founder and CEO, Paradigm
LinkedIn: https://www.linkedin.com/in/joelle-emerson-4a24225
Website: https://www.paradigmiq.com/

Host
William Tincup
Co-founder, WRKdefined
LinkedIn: https://www.linkedin.com/in/tincup

Follow WRKdefined
LinkedIn: https://www.linkedin.com/company/wrkdefined/
Website: https://wrkdefined.com/
Newsletter: https://wrkdefined.com/about-wrkdefined
Apple Podcasts: https://podcasts.apple.com/us/channel/wrkdefined/id6482847601

Powered by the WRKdefined Podcast Network.