Product Management to AI Evals—honestly, it still feels strange even saying these words out loud because I never planned any of this. People imagine career changes happen after some dramatic meltdown or some inspiring life moment… but mine didn’t. Mine started in a slow, almost boring way. Small frustrations here and there, a weird restlessness that kept growing, and a random curiosity that I didn’t take seriously at first.
To be honest, if someone told me three years ago that I would leave Product Management and dive into AI evals of all things… I would’ve laughed. Or maybe ignored them. But life has this funny way of pushing you gently until one day you realize, “Wait, I’m not even the same person anymore.”
Introduction
For years, I followed the typical PM routine: sprint planning, stand-ups, customer calls, backlog grooming, and documents that always needed “just one more revision.” You know that loop, right? It gets comfortable, almost too comfortable. And that comfort becomes a trap before you even notice it.
I wasn’t miserable. But something inside me kept saying, “You’re not excited… this can’t be it.”
And honestly, I kept pushing that feeling away.
Everything changed because of one late-night curiosity spiral. And that tiny moment is what pulled me from Product Management to AI Evals without me even knowing it.
How the Shift from Product Management to AI Evals Quietly Started
There wasn’t some movie-style turning point. It was more like a slow drift that I didn’t recognise until much later.
That one night that didn’t feel important at all
One random evening, I was supposed to finish a requirements doc. Instead, I got distracted—not proudly admitting this—and started reading a Stanford paper about how LLMs are benchmarked.
Half of it felt too technical, and half of it felt way above my head.
But something in it… I don’t know… it sparked something.
I kept thinking, “Why do these models behave like this? How do we check if they’re thinking properly? Does this even count as thinking?”
What was meant to be a 10-minute break became a two-hour deep dive. And, weirdly, I didn’t feel tired at all.
More Info: OpenAI
Losing interest in the usual PM loop
After that, everything in product management felt heavier than usual.
Meetings felt like déjà vu.
Documentation felt like punching keys rather than building ideas.
Customer calls felt like reruns of old episodes.
Some people think PM life is glamorous, but to be real… Most days it’s a mix of juggling expectations and solving problems that feel a bit repetitive after a point.
I started noticing that I wasn’t showing the same enthusiasm I once had. Not for features, not for roadmaps, not for metrics.
And that was my first quiet warning sign.
More Info: Anthropic Model Evaluation
When AI Evals Started Taking Over (Without Asking Me)
This part still feels funny to me because I didn’t “choose” AI evals in a conscious way. It just kept pulling me in.
The weird satisfaction of evaluating how a model thinks
I began experimenting with small tests.
Simple prompts at first.
Then more complex ones.
Then I started comparing outputs like some kind of detective, trying to see patterns or behaviors that weren’t obvious.
It felt like solving puzzles.
And I’ve always been a puzzle person.
Before I realized it, I was staying up late—not for a product release, not for a deadline—but because I genuinely wanted to see how a model would respond if I changed one little instruction.
Curiosity became a full obsession
There were days when I would think, “Let me test one more prompt.”
That one prompt became five.
Then ten.
And suddenly it was 1:30 am.
I wasn’t forcing myself.
I wasn’t dragging myself.
I was pulled in fully—mind, energy, and excitement.
That’s when I understood that my shift from Product Management to AI Evals wasn’t some random distraction. It was a real transition happening inside me.
More Info: Stanford HELM
Why the Transition Made Sense in the Long Run
Looking back, it actually feels like AI evals were a natural continuation of everything I loved about PM—minus the parts that exhausted me.
Structured thinking works in both fields
PMs break problems into small parts.
Eval engineers break model behavior into small parts.
It’s surprisingly similar.
Asking “why” is basically the core skill
In PM, you keep asking why customers behave a certain way.
In AI evals, you keep asking why models behave a certain way.
Different subjects, same mindset.
Also Read: Stanford Just Killed Prompt Engineering With 8 Words (And I Can’t Believe It…)
AI gave me the excitement I didn’t know I was missing
Every time I tested something new, I felt alive. Curious. Engaged.
It wasn’t just work—it was discovery.
That’s not something PM gave me anymore.
Key Points
- My move from Product Management to AI Evals was slow, unplanned, and started with tiny curiosities
- PM routines became mentally repetitive, which I ignored for too long
- AI evals triggered genuine excitement—something I hadn’t felt in years
- Evaluating model behaviour felt like solving living puzzles
- PM skills quietly helped me transition smoothly into the AI world
Also Read: Top AI Tools for Digital Marketing in 2025 (Real Tools That Actually Work)
Conclusion
If someone asked me how I made this switch, I would honestly say, “I didn’t. It happened to me.”
Sometimes you don’t realize you’re outgrowing something until something new pulls you in. And AI evals did exactly that. It gave me the spark I had been missing, maybe even avoiding.
I let go of product management not because it was bad, but because something better—something more ‘me’—appeared.
Final Verdict
If your career feels too predictable, too quiet, or too drained of excitement, don’t ignore the tiny curiosities. Those small things often lead to the biggest shifts.
My journey from Product Management to AI Evals taught me that unexpected paths sometimes become the best ones.
Key Takeaways
- Pay attention to small curiosities
- Excitement is a better guide than comfort
- PM skills translate well into AI evals
- It’s okay to outgrow a career
- The right work will naturally pull you in
FAQs
1. Was switching from Product Management to AI evals scary?
A little, yes. Any change feels scary. But the excitement made it worth it.
2. Do I need deep technical knowledge?
Not at the start. You can grow into it slowly.
3. Is AI evals a stable career?
More stable than people realise. Companies need evaluation experts more than ever.
4. Does PM experience actually help?
Absolutely. The analytical mindset translates beautifully.
Chandra Mohan Ikkurthi is a tech enthusiast, digital media creator, and founder of InfoStreamly — a platform that simplifies complex topics in technology, business, AI, and innovation. With a passion for sharing knowledge in clear and simple words, he helps readers stay updated with the latest trends shaping our digital world.
