The WorkOps Podcast

Summary
Is AI a skill to list on a job description, or simply the way work gets done now? In this episode of The WorkOps Podcast, host Jeet Mukerji sits down with Amy Garefis, Chief People Officer at ZipRecruiter, who spent most of her career as a CPA before moving into HR four years ago. Amy explains why the hardest part of that move was accepting that people decisions rarely have a right answer, and why ZipRecruiter keeps investing in interns and early career talent while other companies pull back. She walks through how her people team is adopting AI, from baseline courses to self-identified evangelists building workflows, and why freed-up time disappears unless you plan for it. She also shares the story of ZipRecruiter's quarterly collaboration weeks in Phoenix, where the original goal of productivity gave way to something the company values more: connection. A practical conversation for people leaders working out how to bring their teams along on AI without losing the human side of the work.

Chapters
00:00 Cold open
00:43 Welcome Amy Garefis of ZipRecruiter
01:06 From CPA to chief people officer
04:11 Why HR decisions rarely have a right answer
05:47 Should new grads start in HR?
08:04 Why ZipRecruiter still hires early career talent
11:11 Is AI turning everyone into a manager?
15:11 Tokens, time and how the people team uses AI
20:38 Building AI fluency across the company
25:08 ZipRecruiter's quarterly collaboration weeks
34:38 Advice for leaders driving AI adoption

Takeaways
- AI is a way of working, not a skill. Amy expects it to appear in job descriptions for now, then fade into a baseline expectation the way "internet" once did.
- HR rarely has a right answer. Moving from accounting meant getting comfortable with decisions everyone can relate to and has an opinion on.
- Automation does not create free time by itself. Decide in advance which strategic work the saved hours will go to, or other tasks will absorb them.
- Make AI adoption accessible. Offer a finite set of tools, share small real examples, and let self-identified evangelists lead inside each department.
- In-person time pays off in connection more than output. ZipRecruiter's collaboration weeks started as a productivity play and became a team building one.

Connect with the Guest
Amy Garefis LinkedIn: https://www.linkedin.com/in/amy-garefis/
Company Website: https://www.ziprecruiter.com/

Sponsor
This episode is brought to you by Kinfolk, the AI service desk built for HR.

See more at kinfolkhq.com

Heather Oxley, CHRO at Perficient, joins Jeet Mukherjee on The WorkOps Podcast to explain how a 7,000 person IT services company is rebuilding its People function around AI without pretending everything can move at the same pace. She opens with the work study she ran on arrival, which found that 60% of HR's workload was inquiry, and shows how that single number set her first priority.

From there the conversation moves through Perficient's two speed operating model, the AI academy whose capstone requires a real client or internal outcome, and the decision to break 40 semi autonomous units into global practices so hiring draws from the largest pool instead of the smallest. Heather is direct about what does not work, including concentrating AI accountability in one person and treating course completion as proof of skill. HR leaders, People Ops practitioners and anyone building an AI operating plan for the function will find a practical sequence here: measure, separate fast from slow, prove the skill, fix the structure, then buy for speed.

Chapters
00:00 Cold open
00:54 Welcome Heather Oxley of Perficient
01:12 From IT consulting to the CHRO seat at Perficient
03:18 Why the CHRO job keeps expanding
05:11 Where HR ends and IT begins
07:06 Spreading AI accountability across the HR leadership team
09:41 The academy, the capstone, and applied AI skill
12:16 Go fast on the obvious, slow on the complex
13:02 60% of HR workload was inquiry
15:26 Build versus buy on hiring and skills
19:29 Global practices and matching supply to demand
23:01 HR in three years, agent supported and proactive
26:20 Pilots, passengers, and leaning into the hard

Takeaways
- A work study at Perficient found that 60% of HR's workload was inquiry, questions already answered somewhere in the policy library, which made an inquiry agent the obvious first place to move fast.
- Perficient runs at two speeds by design: fast where the ROI is clear and the blocker is data hygiene, slower where cross silo handoffs offer a real chance to redesign the employee experience.
- Course completion does not count as capability at Perficient, because the AI academy only certifies people once they finish a capstone applied to a real client or internal situation and assessed on the outcome.
- Rather than appointing an AI transformation lead, Heather gives every function head accountability for AI in their area, on the logic that concentrating ownership in one person slows the whole organization down.
- Consolidating roughly 40 autonomous units into global practices widened the hiring funnel from the smallest pod to the largest, and a bought skills platform, kept current by feeds from learning and daily work, powers the supply to demand match.

Connect with the Guest
Heather Oxley LinkedIn: https://www.linkedin.com/in/heatheroxley/
Company Website: https://www.perficient.com

Sponsor
This episode is brought to you by Kinfolk, the AI service desk built for HR.

See more at kinfolkhq.com

Summary
What happens when a CHRO treats AI like a brainstorming wall instead of a threat? In this episode of The WorkOps Podcast, host Jeet Mukherjee sits down with Jason Desentz, Chief Human Resource Officer at Toshiba America, to unpack why AI is an enhancement story, not a replacement story. Jason shares the history lesson behind his optimism, from direct deposit to factory robots to Ford's reversed AI layoffs, then goes inside Toshiba's HR Shark Tank, where cross functional teams built working Copilot agents like Payroll Princess and Time Tamer in 60 days. He also lays out his two goalpost framework for every AI decision and warns about the integration trap that could fragment HR tech stacks all over again. A practical conversation for HR and operations leaders who want to experiment with AI without disrupting the business.

Chapters
00:00 Introduction
01:45 From police academy to CHRO
05:50 Why AI will deepen HR expertise instead of replacing it
08:50 The direct deposit lesson
13:55 Starting small with proof of concept
16:35 Inside Toshiba's HR Shark Tank
20:00 Build, borrow, or bot
21:30 The two goalposts and the integration trap
28:55 FOBO and getting people AI ready
32:55 From curious to cautious

Takeaways
- AI will change how HR work looks, not whether it exists. Like computers, direct deposit, and factory robots before it, it enhances the function and creates new work.
- Start small instead of trying to do everything at once. Pick one proof of concept and give it a full cycle, a year to 18 months, before judging whether it worked.
- Cross functional experimentation multiplies value. Toshiba's HR Shark Tank mixed payroll, field HR, business partners, and L&D to build real -Copilot agents on top of their day jobs.
- Every AI decision has to pass two goalposts: little to zero disruption to the business and a genuine improvement to the employee experience.
- Watch the integration trap. HR spent 15 years consolidating eight systems down to three, and bolting on AI point solutions risks recreating the same fragmentation.

Connect with the Guest
Jason Desentz LinkedIn: https://www.linkedin.com/in/jason-desentz/
Company Website: https://www.toshiba.com/

Sponsor
This episode is brought to you by Kinfolk, the AI service desk built for HR.

See more at kinfolkhq.com

Summary
Payroll problems were blowing up Slack channels at SurveyMonkey, and employees weren't taking them to finance. In this episode of The WorkOps Podcast, host Jeet Mukerji talks with Becky Cantieri, Chief People Officer at SurveyMonkey, about the decision to hit the blow-up button and move payroll out of finance and into HR, how a willing CFO and one simple question (who is willing to talk to the employee?) made the handoff a slam dunk, and why the defining skill for HR leaders right now is metabolizing change: finding signal in noise fast enough to prioritize, pivot, and keep up in the AI era. A practical listen for people leaders, HR operations teams, and anyone rethinking who should own the employee experience.

Chapters
00:00 Cold open
01:16 From the Nordstrom sales floor to CPO
06:37 Fifteen years of reinvention
09:26 Metabolizing change
14:22 The payroll problem
18:31 Winning over the CFO
21:44 Rebuilding the team
28:03 The AI champion at SurveyMonkey
32:51 Refusing to wait in the AI queue
35:21 Permission, capability, agency

Takeaways
- Employees don't think in org charts. When something breaks, they go to whoever usually answers, so design ownership around the employee experience, not department history.
- The team willing to talk to employees should own the employee-facing function. SurveyMonkey moved payroll into HR when it became clear finance didn't want those conversations.
- Make handoffs safe with data: SurveyMonkey's HR team committed to SLAs and email volume reporting while banking controls stayed with finance.
- The defining HR skill right now is metabolizing change: finding signal in noise fast enough to prioritize, pivot, and keep up with AI.
- Metabolizing is teachable. It starts with curiosity, a growth mindset, and the willingness to throw out a process you built and are proud of.

Connect with the Guest
Becky Cantieri LinkedIn: https://www.linkedin.com/in/rebecca-cantieri-85654a/
Company Website: https://www.surveymonkey.com

Sponsor

Summary
What happens when a CEO hands unlimited AI access to every employee, and adoption still stalls? In this episode of The WorkOps Podcast, host Jeet Mukerji sits down with Kimberly Nerpouni, Global VP of People & International Operations at Pearl, the parent company of JustAnswer. Kimberly shares why AI transformation is a human enablement problem rather than a technology problem, how her People Ops team built Pearl's AI accelerator and a 1 to 10 adoption scale with no bad scores, and her three levels of work framework for deciding what AI should absorb first. She also explains the feedback facilitator agent her team is building, the hard line she draws (AI never gives feedback, it helps managers facilitate it), and why PIPs don't exist at Pearl. This conversation is for people leaders, HR teams, and anyone navigating AI adoption inside their organization.

Chapters
00:00 Introduction
00:45 From IT manager to people operations
04:25 JustAnswer, Pearl, and human plus AI
06:15 Inside Pearl's AI accelerator
10:55 Build, buy, or borrow
13:15 The three levels of work
17:05 AI that facilitates feedback instead of giving it
19:50 The AI agent hub and transparency
23:35 Why PIPs don't work and what replaces them
35:05 Don't wait, just do

Takeaways
- AI adoption is a human enablement problem, not a technology problem, and people ops is uniquely positioned to lead it.
- Move AI into level one work first, the tasks that don't require your expertise, so your team can focus on level two and level three work.
- Build, buy, or borrow: piloting with AI startups can reveal exactly what's worth building bespoke in-house.
- AI should facilitate feedback conversations by scanning context and prompting managers, but it should never give the feedback itself.
- Front-load clarity with job descriptions, career ladders, and employee-owned development plans so PIPs are never needed.

Connect with the Guest
Kimberly Nerpouni LinkedIn: https://www.linkedin.com/in/kimberlypignolet/
Company Website: https://www.pearl.com

Sponsor
This episode is brought to you by Kinfolk, the AI service desk built for HR.

See more at kinfolkhq.com

Summary
What happens when your AI pilot works too well to stay a pilot? In this episode of The WorkOps Podcast, host Jeet Mukerji talks with Ivan Nosov, Head of HR Tech and Global Total Rewards Director at Campari Group, about taking AI in HR from proof of concept to production. Ivan shares why he started with Copilot Studio knowing it wouldn't last, where RAG breaks down at enterprise scale, and how his team built a digital twin router app on Claude that loads only the context each question needs. He also lays out the adoption playbook that made it stick: winning over regional directors and COE champions first, running hackathons for awareness, and addressing the work people hate instead of the work they love. A practical conversation for HR, people ops, and workplace technology leaders navigating AI transformation.

Chapters
00:00 Introduction
01:30 From IT engineer to HR leader
03:45 Should AI transformation sit in HR or IT
05:55 How non technical HR can get started
07:20 Copilot Studio as a proof of concept tool
09:30 Harness, routing, and the limits of RAG
15:00 Building the digital twin router app on Claude
17:40 Context engineering and distilling tacit knowledge
19:30 Winning adoption through champions
29:55 The future of junior roles and build versus buy

Takeaways
- Start with the accessible tool to prove the concept, then move on. Copilot Studio validated Campari Group's AI ideas, but production required control over the harness that abstracted platforms can't offer.
- Routing beats RAG at scale. Loading only the relevant context for each question makes AI more targeted, more efficient, and far less likely to hallucinate.
- Campari Group's digital twin runs on Claude with deliberate model selection: Sonnet for efficiency, Haiku for helper queries, and Opus for final artifacts because of its design taste.
- Adoption spreads through champions. Win over regional directors and COE heads first, stay with them until the outputs click, and they will share it further than any rollout plan.
- Context engineering is the real work. Distill tacit knowledge into a compact, unambiguous knowledge base: 250,000 tokens in total, with no session loading more than 50,000.

Connect with the Guest
Ivan Nosov LinkedIn: https://www.linkedin.com/in/ivan-nosov/
Company Website: https://www.camparigroup.com/

Sponsor
This episode is brought to you by Kinfolk, the AI service desk built for HR.

See more at kinfolkhq.com

Summary
What happens to careers when AI makes more than half of all roles look fundamentally different within two years? In this episode of The WorkOps Podcast, host Jeet Mukergi sits down with Lisa Sherwell, Chief People Officer at SUSE, to unpack her answer: strive for job security, not role security. Lisa shares how SUSE's Future Selves program transformed engagement for its over 50 workforce, why internal gigs are the new engine of career growth, and why performance management fails not because of broken processes but because leaders avoid honest conversations. A practical, candid conversation for HR leaders, people operations professionals, and executives navigating the AI era.

Chapters
00:00 Introduction
01:15 From sales and call centers to chief people officer
02:30 Building commercial curiosity in HR
04:15 Future Selves: investing in the over 50 workforce
07:00 Job security vs role security
08:55 Hiring for AI potential, not expertise
10:30 How SUSE rolled out AI across the business
16:45 Running HR for 8% less while delivering more
24:30 Performance is a clarity problem, not a process problem
33:00 A new playbook for the AI era

Takeaways
- Strive for job security, not role security: broad competencies and internal mobility protect people as AI reshapes their roles.
- More than half of all roles could look fundamentally different within two years, so career agility is now a core business responsibility.
- SUSE's Future Selves program proved that investing in the over 50 workforce lifts engagement, internal mobility, and community, while preserving invaluable business context.
- Performance is a clarity problem, not a process problem: honest meets, fails, exceeds conversations beat new ratings scales and OKR frameworks.
- Ownership of performance belongs with the individual contributor, because nobody cares more about your career growth than you do.

Connect with the Guest
Lisa Sherwell LinkedIn: https://www.linkedin.com/in/lisasherwell/
Company Website: https://www.suse.com

Sponsor
This episode is brought to you by Kinfolk, the AI service desk built for HR.

See more at kinfolkhq.com

Summary
On this episode of The WorkOps Podcast, host Jeet Mukergi sits down with Natasha Valani, Chief People Officer at Tubi, the number one free ad-supported streaming service in the US. Natasha shares how her people team built a live calibration tool in Replit in just two days, connecting it back to Workday and saving roughly two weeks during year end review season. The conversation goes well beyond one tool: Natasha unpacks how she hires for high agency and low ego, why subtraction is her favorite leadership discipline, and how she sees junior and people roles evolving as AI takes on more of the routine work. It's a practical, candid look at running a lean, builder-minded HR function, ideal for people leaders, HR practitioners, and anyone rethinking how their team operates in the AI age.

Chapters
00:45 Welcome and Natasha's story
01:45 From 26 schools to belonging and the people space
03:45 Building a culture where everyone builds
06:45 Hiring for high agency and low ego
09:45 The year end calibration problem
10:45 The two day tool built in Replit
15:45 Why build instead of buy
19:45 Trade-offs and the power of subtraction
21:45 The honest gap, headcount management
26:45 The future of junior roles in the AI age

Takeaways
- You don't need a vendor RFP to solve an HR problem. Someone who knows the guts of the process built a live calibration tool in Replit in two days and connected it to Workday, saving roughly two weeks.
- High agency is the differentiator in the AI age. Natasha screens for it with one question: tell me about a time you built a solution nobody asked you to.
- Pair high agency with low ego. Stay curious and ask questions, but don't assume nothing good existed before you.
- Subtraction is a strategy. Force-rank priorities, openly showcase what you removed, and protect your team's bandwidth by refusing to do everything.
- Junior roles won't disappear, they'll shift. The people closest to the guts of a workflow become operational advisors and orchestrators, and AI still needs a human eye to catch the slop.

Connect with the Guest
Natasha Valani LinkedIn: https://www.linkedin.com/in/natasha-valani-1758a418/
Company Website: https://tubitv.com

Sponsor
This episode is brought to you by Kinfolk, the AI service desk built for HR.

See more at kinfolkhq.com

Summary
Carmel Smith, Director of People Operations and Programs at Tenstorrent, joins host Jeet Mukerji on The WorkOps Podcast to make the case that the AI moment is really a human one. Fresh off building people operations inside a company that grew headcount 43% past 1,300 employees, Carmel explains why she's rebranding her team from "people operations" to system architects and experience creators, why she refuses to mandate AI or track tokens and agents, and what she means when she says she's leading a human revolution rather than an AI one. It's a refreshingly optimistic, practical conversation for HR and people leaders, operations teams, and anyone trying to bring humans along through rapid change.

Chapters
00:00 Cold open, the scary part of getting AI right
00:45 Meet Carmel Smith and Tenstorrent
03:45 From people operations to system architects
06:45 Why she never mandates AI
07:45 The painting lesson and psychological safety
11:45 Leading a human revolution
12:45 People data as an open source foundation
20:45 Redefining productivity beyond AI ROI
30:45 The monthly hackathon her leader defends

Takeaways
- HR's real identity shift is internal: help your team see themselves as system architects and experience creators, not process executors, and their confidence follows.
- Mandating AI backfires. Counting tokens, hours, and agents produces box checking, not rethinking. Curiosity, safety, and a genuinely excited leader scale far better.
- AI rewards the most human skills. Having strong opinions and articulating them clearly matters more than technical knowledge, because the systems part is the easy part.
- Don't track ROI on AI directly. If it's implemented well it bleeds into everything, so measure faster processes, redesigned workflows, and reduced fear instead.
- Protect creative time on purpose. A defended monthly hackathon, even when the output fails, changes how a team works by redefining what productivity means.

Connect with the Guest
Carmel Smith LinkedIn: https://www.linkedin.com/in/carmelmoyal/
Company Website: https://tenstorrent.com
Disclaimer: The views and opinions expressed here are my own and do not reflect the official policy or position of Tenstorrent.

Sponsor
This episode is brought to you by Kinfolk, the AI service desk built for HR.

See more at kinfolkhq.com

Summary
On this episode of the WorkOps Podcast, host Jeet Mukergi talks with Kate Stewart, Staff People Operations, AI and Automations Lead at Horizon3.ai, about what really happens when you put an AI agent into a live people-operations workflow. Kate shares the story of the offer-validation agent she built in Slack to check every job offer against approved comp bands and job architecture, how a Slack update broke it after two months, and the two weeks she spent secretly becoming the agent herself to keep the quality bar from slipping. Along the way she unpacks why a hallucinating agent is more dangerous than a silent one, why automation raises the bar instead of lowering it, and how to ship, break, and maintain automations without burning out. It's a candid, practical listen for anyone in people ops, HR tech, or operations moving from curiosity about AI to actually running it in production.

Chapters
00:00 From the retail floor to people ops
06:30 Staying curious without a second job
07:45 The broken process behind every offer
09:45 Building the offer validator in slack
11:45 Two months in, the agent breaks
12:45 Becoming the human agent
13:45 Why a hallucinating agent is a liability
14:55 Automation raises the bar
15:45 From perfectionism to shipping V1
28:45 Build it, let it break, protect the time

Takeaways
- When an automation breaks, you don't fall back to your old baseline, you fall below it, because automation raises the bar for what your team considers acceptable
- A silent agent is confusing, but a hallucinating agent is a liability, a confident wrong answer is far more dangerous than no answer at all
- Job architecture is not a data hygiene problem, it is an IT provisioning problem, the wrong title in the HRIS means the wrong system access on day one
- Build one agent to do one job at the right moment, rather than spreading the same check across three people and three separate touchpoints
- Ship V1, let it break, and protect the time to maintain it, the build teaches you the problem and the break teaches you what you actually built

Connect with the Guest
Kate Stewart LinkedIn: https://www.linkedin.com/in/kate-stewart00/
Company Website: https://www.horizon3.ai

Sponsor
This episode is brought to you by Kinfolk, the AI service desk built for HR.

See more at kinfolkhq.com