​Go listen to the latest episode of The Modern People Leader — where we sat down with Diane Sadowski-Joseph, Co-Founder and Head of Product at Clarinet (formerly ApplyAI), to talk about how HR teams can actually move from AI curiosity to AI execution.

If you’ve ever felt behind on AI — or overwhelmed by the sheer volume of tools, workflows, and “10x” case studies flooding your feed — this one’s for you.

Here’s what stood out most from the conversation. ⬇️

“Build AI on something — not something on AI.” — Diane Sadowski-Joseph

1. You’re not behind — you’re comparing outputs, not inputs

It’s easy to scroll LinkedIn and feel like everyone else has already automated their entire function. But Diane made a critical distinction: most teams are comparing outcomes without understanding the inputs required to get there.

For AI to truly unlock value, you need three things:

• A somewhat standardized process
• Clean, accessible data
• A consistent problem + consistent output

Not many.

AI is often marketed as plug-and-play magic. But in reality, it thrives in structured environments. If your workflows are chaotic, your data is messy, and your output constantly changes, AI won’t save you — it will amplify the mess.

The takeaway: before chasing the latest model release, ask whether your inputs support the outcome you’re expecting.

🏡 A better way to do backup care

Backup care has felt especially real for us lately. With Stephen navigating life with a newborn plus two teenage daughters, and Daniel raising a toddler - we find ourselves dealing with sick days, last-minute gaps, and those moments when childcare plans fall apart at the worst possible time. It’s also when you realize quickly whether your backup care benefit actually works… or just looks good on paper.

That’s the issue a lot of employers are running into these days. Traditional backup care wasn’t built for how families live and work today - so when employees need it most, it doesn’t show up.

That’s why more companies are switching to Wellthy Backup Care.

Here’s what that looks like in practice:

• Flexible backup care for kids, adult dependents, and even pets
• Access to a vetted provider network or a Wellthy expert who handles booking
• Pay-as-you-go pricing so you only pay for care that’s actually used
• Onsite and nearsite options for in-person teams

Backup care has changed. See the difference with Wellthy.

👉 Get a personalized geoaccess report​

2. Stop chasing the latest model. Solve today’s problem.

One of the most freeing insights from Diane: the newest AI release isn’t necessarily the one that solves your problem.

It’s easy to fall into the trap of constantly upgrading — new models, new agents, new copilots. But most organizations haven’t fully implemented what’s already available.

Instead of asking:
What’s the most cutting-edge tool?

Ask:
What business pain are we trying to solve right now?

Diane framed AI adoption more like kaizen — continuous improvement — rather than a one-time “AI implementation.” Your workflow today will evolve in three months. That’s not failure. That’s iteration.

The use case implemented today beats the perfect use case you’re still designing.

3. If you don’t know where to start, automate work chores

Diane introduced a simple prioritization lens: the Five Frictions framework. But if you’re short on time, here’s the shortcut:

Start with work chores.

These are the time-consuming, low-leverage tasks that:

• Drain cognitive energy
• Don’t require deep strategic thinking
• Often fall into the “rubber ball” category (drop it, and it bounces)

Work chores are powerful because:

• The ROI is obvious (time + energy savings)
• The risk is lower than experimenting on mission-critical systems
• They build momentum quickly

Instead of trying to “AI-ify” your entire function, find the repetitive tasks that make you sigh every time you open them.

That’s your starting point.

📊Turn your people data into usable insight

In our MPL webinar on how HR teams are actually using AI, leaders like Jenny Molyneaux (Vercel) and Brandon Sammut (Zapier) kept highlighting the same idea: AI only matters if it makes insights usable. Teams don’t need more dashboards - they need clarity.

That’s exactly what Ask ChartHop is built for. You type a question and instantly get an answer. When we saw it at HR Tech, people were surprised by how fast it worked. Org charts, comp bands, churn risk, hiring timelines, pay equity views, headcount plans - all generated in seconds. No spreadsheets, no BI requests, no “I’ll get back to you.”

Here’s what that looks like in practice:

• Shareable charts and clear visuals
• Sortable lists you can act on immediately
• Access controls that protect sensitive data
• Insights leaders can use to make better decisions

It turns data into usable insight.

​👉 Try Ask ChartHop​

4. The best AI use cases are tied to real KPIs

One of Diane’s client examples stood out: a fast-scaling fintech company trying to reduce time-to-hire.

Rather than brainstorming random AI ideas, they anchored everything to a business goal.

They realized hiring managers were slowing down candidate feedback. So they built a simple GPT that:

• Took interview transcripts
• Allowed messy bullet points from hiring managers
• Transformed them into high-quality, legally compliant candidate feedback
• Drafted the email for recruiting to send

What used to take days of shoulder-tapping now took minutes.

The result:

• Faster time-to-hire
• Higher candidate experience
• Less friction between recruiting and hiring managers

That line stuck with us.

If you’re not tying AI to a KPI you’re already accountable for, you’re probably building theater — not leverage.

5. Click-cutters are the compounding interest of AI

Some of the most powerful wins aren’t flashy agents or dashboards. They’re what Diane calls “click cutters.”

These are tiny workflow improvements:

• Automatically converting transcripts to Google Docs
• Auto-filing documents into the right folders
• Reducing copy/paste steps
• Eliminating unnecessary tab switching

It’s not just about saving time — it’s about preserving mental energy. The more micro-frictions you eliminate, the more strategic capacity you create.

Click cutters compound.

Individually, they seem small. Collectively, they change how your day feels.

💙 Make leave feel human - not chaotic

In one of our recent webinars, we talked about how leave is usually put into three buckets: compliance and pay, business continuity, and the very real human experience happening at home.

This couldn’t have hit harder for me and Stephen. We both recently became parents - and it showed how complex leave can be, even on a tiny team.

That’s exactly what Tilt is built for.

Tilt is a leave experience platform that manages all types of leave, nationwide - from FMLA and PFL to everything in between - while giving HR, managers, and employees clear visibility into what’s happening. It combines compliant leave management with a genuinely human experience.

Here’s what that looks like in practice:

• Personalized leave plans for every employee
• Role-based visibility for HR, managers, and employees
• AI that automates admin and answers routine questions
• Compliance-backed, expert support

It turns leave from paperwork into a people-first leave experience.

👉 Learn more about Tilt​

6. Bring the tech to the people — not the people to the tech

One of the most practical shifts we discussed: stop trying to force behavior change if you don’t have to.

Instead of:
“Go to Confluence to find the policy.”

Ask:
“How do we bring the policy to where people already work?”

That might mean:

• HRIS integrations inside Slack
• FAQ bots inside your primary communication tool
• Personalized dashboards pushed proactively to execs

AI enables hyper-personalized, in-context experiences. And when you reduce the friction between people and information, adoption skyrockets.

Thanks for reading. See you next time!

P.S. If you like MPL, help us grow the show by giving us a 5 star rating on Apple or Spotify. ⭐⭐⭐⭐⭐

​​​

600 1st Ave, Ste 330 PMB 92768, Seattle, WA 98104-2246
​Unsubscribe · Preferences​