AI Summary

• Parenting young children reveals that effective leadership prioritizes understanding over expertise and creates psychological safety for innovation
• The most powerful leadership tool is genuine curiosity - asking “why” like a 4-year-old uncovers breakthrough solutions hidden beneath assumptions
• Building team capability through shared learning creates stronger, more resilient organizations than relying on external expertise
• Creating buffer zones between organizational chaos and team execution requires the same patience and consistency needed in parenting

The Accidental Leadership School

I thought I had leadership figured out. Read all the right books, attended the workshops, earned my MS, managed data science teams at a large utility for years. Then I became a dad and realized I knew absolutely nothing.

My 1-year-old daughter and 4-year-old son don’t care about my degrees or how many machine learning models my teams have deployed. They just want me to understand them, be curious with them, and not lose my mind when things go sideways.

Turns out, that’s exactly what my data science teams needed too.

After four years of fumbling through both parenting and leadership simultaneously, I’ve discovered that toddlers make surprisingly good leadership coaches. They’ve completely changed how I think about empathy, curiosity, and creating safe spaces for people to grow - especially when you’re trying to shield your teams from the chaos swirling around the C-suite.

Actually Listen Before You Try to Fix Everything

My 1-year-old can’t tell me what’s wrong when she’s upset. She just cries. So I’ve become a detective - is she hungry? Tired? Frustrated that she can’t reach her toy? I have to put aside whatever I was doing and really pay attention to understand what she needs.

This completely changed how I handle team problems.

Before kids: Someone comes to me with an issue about model performance, I immediately start debugging the algorithm.

After kids: I force myself to understand first, solve second.

Instead of jumping to technical solutions, I started asking things like:

“Help me understand what you’re experiencing day-to-day with this model” “What’s the most frustrating part of this deployment process?” “Walk me through how this data pipeline failure impacts your work”

One machine learning engineer came to me complaining about “drift in our energy forecasting models.” My old approach would have been to immediately dive into retraining strategies and feature engineering. Instead, I asked her to help me understand what she was experiencing.

The real issue wasn’t model drift - it was that constant organizational restructuring meant she never knew if her project would still exist next quarter. She couldn’t focus on the technical work because she was stressed about job security. Completely different problem, completely different solution.

Now when senior leadership announces another strategic pivot, I spend more time understanding how my team is processing the change than I do diving into the technical implications.

Ask “Why” Like Your Job Depends on It

My 4-year-old asks “why” about everything. And I mean everything.

“Why is the sky blue?” “Why do cars have wheels?” “Why do we need to eat vegetables?”

It’s exhausting, but it’s also brilliant. He doesn’t accept “that’s just how things are” as an answer.

I started bringing that same energy to work, and it’s transformed how we approach problems in the utility space.

In strategy meetings, I became the person asking:

“Why do we forecast energy demand this way?” “Why do we segment customers using these particular features?” “Why do we assume this regulatory requirement can’t be automated?”

At first, people were a bit put off. These questions can feel naive when you’re deep in domain expertise. But they’ve led to some of our biggest breakthroughs.

One simple “why do we retrain this model monthly instead of weekly?” led us to completely redesign our MLOps pipeline and improve our peak demand forecasting accuracy significantly. Sometimes the most basic questions unlock the biggest improvements.

The best insights often hide behind the assumptions we’ve stopped questioning.

I now schedule “fundamental questioning” sessions quarterly where we deliberately challenge our basic assumptions about how we approach machine learning in the utility sector. The only rule: every “why” gets a real answer, not defensiveness.

Learn Stuff Together (Even When You Don’t Know the Answer)

Watching my daughter discover shadows for the first time, or my son realize he can mix colors to make new ones - the pure joy of discovery is infectious.

And here’s the thing: I’m learning right alongside them. I don’t have all the answers about why leaves change color or how airplanes stay up. We figure it out together.

This completely changed how I approach new challenges at work.

Instead of feeling like I need to be the expert on everything from transformer models to grid optimization, I started saying: “I don’t know, let’s figure it out together.”

When we needed to implement computer vision for infrastructure inspection, instead of bringing in expensive consultants, I formed a learning team. I participated not as the director with all the answers, but as someone equally curious about how this technology could work in our specific context.

The project took longer than it might have with outside experts, but something amazing happened - the entire team’s confidence and capability grew exponentially. They owned the solution because they built it together. More importantly, they could troubleshoot and improve it without depending on external resources.

The best leaders create environments where everyone gets to discover and learn together.

Make It Safe to Mess Up

My son built this elaborate castle out of magnetic tiles last week. Spent 30 minutes on it, was so proud. Then he accidentally knocked it over while reaching for something.

I was ready for a meltdown. Instead, he looked at the pieces and said, “Now I can build something even cooler!” and started over.

That resilience didn’t happen overnight. It developed because we’ve always treated mistakes as learning opportunities, not disasters.

I realized my data science teams needed that same psychological safety, especially when working in an environment where C-suite priorities shift like weather patterns.

I started being more open about my own mistakes:

• Sharing my “learning moments” in team meetings
• Running “failure retrospectives” focused on what we learned, not who screwed up
• Celebrating bold attempts, even when they didn’t work out
• Being transparent about organizational changes while maintaining confidence in our team’s mission

After one particularly visible project failure - a customer segmentation model that completely missed the mark - instead of looking for someone to blame, I guided the team through a learning-focused retrospective. The insights from that session informed our next attempt, which became one of our most successful customer analytics initiatives.

More importantly, people started proposing bolder ideas because they weren’t afraid of being punished for trying something that might not work.

The Shield Between Chaos and Execution

This is where parenting really prepared me for utility leadership. Kids need consistency even when the adult world is chaotic. My 4-year-old doesn’t need to know that I’m stressed about work - he just needs his bedtime routine to happen reliably.

In the same way, my data science teams don’t need to absorb every strategic pivot, reorganization, or C-suite shuffle. They need consistent support, clear priorities, and protection from organizational turbulence so they can focus on building great models and solving real problems.

My job became less about managing up and more about buffering down.

When leadership announces another “transformation initiative,” I translate it into: “Here’s what this means for our current projects, here’s what stays the same, and here’s how we’re going to continue delivering value.”

When priorities shift quarterly, I help the team understand which core capabilities remain constant and which tactical directions might evolve.

The best gift you can give your team isn’t perfect information - it’s consistent support and clear focus.

What Changed (And What It Actually Looks Like)

Before Kids (The “Expert” Approach)

• Jump straight to technical solutions when people bring problems
• Feel pressure to have all the ML answers
• Focus on individual model performance
• Treat failed experiments as wasted resources

After Kids (The “Human” Approach)

• Listen first, understand the real problem, then solve
• Comfortable saying “I don’t know this domain, let’s learn together”
• Focus on team learning and collective capability
• Treat failed experiments as essential learning investments

The Practical Results

Employee satisfaction metrics around trust in leadership and engagement showed strong improvements after we implemented the “understand first” approach.

Innovation attempts increased significantly after we created psychological safety around experimentation failure.

Team capability grew faster when we learned new ML techniques together instead of always bringing in outside consultants.

Better solutions emerged when we actually understood business problems before trying to optimize algorithms.

The Uncomfortable Truth About Technical Leadership

Here’s what parenting taught me that all the data science leadership content missed:

Technical expertise isn’t what makes great leaders. The ability to create environments where people feel understood, curious, and safe to experiment - that’s what makes great leaders.

The smartest person in the room isn’t always the best leader. Sometimes the best leader is the person asking the most basic questions or admitting they don’t understand the business context.

Innovation requires psychological safety. Data scientists won’t propose creative approaches if they’re afraid of looking stupid or being blamed when experiments don’t yield immediate results.

Real empathy is work. It means putting aside your own technical agenda and really trying to understand someone else’s experience with the tools, processes, and organizational pressures they face daily.

What This Looks Like in Practice

Start Every Problem Conversation with Understanding

• Instead of: “Let’s tune the hyperparameters”
• Try: “Help me understand how this model’s performance is impacting your workflow”

Ask Fundamental Questions Regularly

• Instead of: Accepting “that’s standard practice in utilities”
• Try: “Why is this the best approach for our specific use case?”

Learn Alongside Your Team

• Instead of: “I’ll bring in a consultant to teach you transformers”
• Try: “I haven’t worked with this architecture either - let’s figure it out together”

Make Experimentation Safe

• Instead of: “Why didn’t this model perform better?”
• Try: “What did we learn from this experiment that will help our next approach?”

So Where Do We Go From Here?

My kids don’t care about my technical credentials or management philosophy. They just want me to pay attention, be curious, and create a safe space for them to grow and explore.

My data science teams need exactly the same thing.

The best technical leaders aren’t the ones with the deepest ML expertise - they’re the ones who can create environments where everyone feels heard, stays curious, and isn’t afraid to try new approaches to solving complex problems.

So if your leadership approach feels stale, maybe the best teachers aren’t in the conference rooms or technical blogs. Maybe they’re the ones asking “why” for the hundredth time today while building something amazing out of blocks.

Four years into this parenting and leadership journey, I’m convinced that the human skills matter way more than the technical ones. My kids keep teaching me that every day - especially when I’m trying to shield them from the chaos while helping them build something beautiful.