I left home at 19 and moved to another country with no money, no safety net, and no plan.
What I had was an inner certainty that I was meant to build something meaningful.

Since then, I have navigated reinvention many times, across leadership roles, entrepreneurship, personal loss, burnout, and rebuilding from scratch. Those experiences shaped how I lead, how I listen, and how I guide others through change.

Today, I work at the intersection of leadership, human dynamics, and artificial intelligence.

I help leaders and organizations navigate transformation without losing trust, clarity, or their humanity.

I am the CEO of APEK.ai, where we build secure, enterprise-grade Agentic AI and voice solutions that modernize operations while humanizing communication. My work focuses on helping organizations adopt AI in ways that feel intelligent, responsible, and deeply human.

With over 20 years in enterprise transformation, leadership strategy, and reinvention, I bring a perspective that blends technology, psychology, and real-world leadership experience.

⸻

Who I Work With

• Executives and boards navigating AI-driven change
• Leaders in transition who look successful on paper but feel misaligned inside
• Organizations ready to evolve beyond legacy thinking and systems
• Founders and senior leaders preparing for their next chapter of impact

People usually come to me when change stops being theoretical and becomes personal.

⸻

What I Believe

AI will not replace leadership.
But it will expose it.

The future belongs to leaders who can combine intelligence with clarity, technology with trust, and innovation with humanity.

If my work resonates, explore my writing, talks, and leadership resources.
Or reach out if you are looking for a trusted advisor, speaker, or strategic partner during moments of transformation.

Eva Karnaukh

I spent 17 years at Google leading engineering teams, 220+ engineers across Ads, Maps, and AI. I loved the work. Building at scale and watching people grow into leaders themselves.

What I'm Building: 

I'm writing Activate: Go Deep. Move Fast. A Leadership System for Builders. 

A book about how leaders build teams that execute quickly and adapt through uncertainty. It's the system I wish I'd had earlier in my career drawn from 17 years of building at Google and interviews with leaders across Netflix, Meta, Datadog, and more.

Alongside the book, I work directly with a small number of leaders navigating rapid growth or transformation, usually through focused sessions or short-term advisory. 

My work focuses on:

- Structural Clarity: Moving past generic alignment to engineering-minded systems that scale.

- Decision Velocity: Helping leaders navigate organizational complexity and AI transformation without stalling.

- Execution Diagnosis: Discover the friction points in your execution model and implementing the "Activate" framework to enable your teams. 

How to Work With Me

- The Newsletter: Join 40,000+ leaders who follow my work, a weekly 3-minute briefing on leadership systems. www.melodyolson.com/newsletter

- Velocity Solution Sessions: A 75-minute diagnostic to unblock your team’s speed and scale your leadership: www.melodyolson.com/advising

- Strategic Advisory: Limited capacity for longer-term partnerships with leaders navigating rapid growth or transformation.

- Speaking: Available for keynotes and workshops on high-velocity leadership and AI transformation.

Melody Olson

Over 17 years at Google, I saw the same pattern again and again: talented people doing strong work, wondering why it wasn't landing. Often, the work wasn't the problem. The visibility was.

A mistake I made early on: assuming good work was enough.

Making your work visible isn’t bragging. People can’t advocate for you in rooms you’re not in if they don’t understand your impact. And people understanding your work is good for you and good for your team. 

A few things to do this better:

1. Communicate simply
→ Share progress in simple, repeatable ways. Send a short weekly update to your manager: what is done, what’s blocked, what’s next.

2. Make your impact clear
→ Don’t just report activity. Frame updates as problem → solution → result. 
Tailor to what each audience cares about.

3. Create touchpoints
→ Introduce yourself early in projects. Create periodic touchpoints with stakeholders. Don't wait until the end.

4. Write it down
→ Share your ideas in writing. Document tradeoffs and outcomes.  

Your work matters.

Make sure the right people can see it.

🔗 I write weekly about engineering leadership and growth. Link in bio: Melody Olson.

Leadership for Product & Engineering | Ex-Google Senior Engineering Director | Activate: Go Deep. Move Fast.

Most teams don't fail from bad ideas.


They fail from chasing too many.

I've seen it over and over, teams full of capable people and good ideas, but no momentum toward what matters most 

Here's what I learned the hard way:

The problem is rarely the ideas. It's the conditions.
→ Too many priorities = no priorities 
→ Ownership without clarity = motion without impact. 

The teams that succeed do less, but they do it with conviction.

1. They kill good ideas to protect great ones
2. They do small tests before they invest
3. They go deep on why this matters before going wide on how to build it
4. They make the hard calls early, not after months of discussion.

The best teams don't just chase ideas.
They build the conditions where the right ideas land.

➕ Follow me, Melody Olson, for what high-velocity teams do differently.

The best leaders I've worked with asked more questions than they gave answers. But that didn't mean they lacked expertise. 

They had plenty. 

Early in my career, I worked with a VP who I found intimidating. He dove deep. Asked hard questions. Pushed my thinking in ways I didn't expect.

I would hesitate and often defer back to him. Then one day in a 1:1, he said to me: "I'm expecting you to be the expert. You're the one spending time on this problem. You own it. I will ask hard questions, but I want you to have an opinion."

That changed how I showed up. As a leader, I've learned: the questions aren't separate from growing your team. They're how you do it.

Image credit: Janis Ozolins 

📌  Go deep. Move fast. I write for engineers and leaders. Link in bio: Melody Olson.

Technical talent matters.
But after working with hundreds of engineers, I've learned something else matters more:

How you show up for your team.

The engineer who gets a lot done but leaves wreckage behind, doesn't help the team move faster.

The one who asks clarifying questions, disagrees respectfully, and makes others better is essential.

But being a great teammate doesn’t mean being agreeable.

Here's what separated the ones who kept advancing from the ones who didn’t:

→ Follow-through. Do what you say you'll do. It is less common than you think.
→ Calm under pressure. Show up well even when there is a fire
→ Ask the question no one's asking. Especially the question others avoid. That person becomes irreplaceable.
→ Disagree without making it personal. Challenge the idea, not the human. This is rare and valuable.
→ Take feedback as data. The teammates who grew fastest treated feedback like data, useful information, not an attack.

Technical skill gets you hired.

How you show up determines how far you go.

🔗 Join my weekly newsletter to learn where to go deep so your teams move faster as things change. https://lnkd.in/gVhWe85t.

Overthinking feels safe. 

But it's fear in disguise.

You can't steer a parked car. But that's what I've watched so many leaders try to do, including me early on. 

Wait for more data. Wait for alignment. Wait until we're "ready."

Meanwhile, the team sits frozen. Momentum slows. The window closes.

Patrick Collison got this right:
"If you can make twice as many decisions at half the precision, that's often better."

Here's the reality:
→ Most decisions are reversible. The cost of being wrong is almost always smaller than the cost of being slow.
→ Momentum creates clarity. You learn more by doing. 
→ Certainty is a myth. Will more analysis actually change the call, or just postpone it? Often you are choosing between two good options. 
→ Waiting isn't free. Your team feels it. They lose energy. The best ones start to wonder why they're stuck.

The leaders I admired most weren't the ones who got it right every time. They decided. Learned and adjusted faster than everyone else.

Speed builds.

Hesitation doesn't.

🔗 Join my weekly newsletter to learn where to go deep so your teams move faster as things change. Link in comments.

The hardest shift in my career wasn't technical.

It was realizing that empowerment requires more work, not less.

I thought I was doing it right. I trusted my team. Gave them autonomy. Dove deep in building systems and stayed out of their way so they could build too.

But I hadn't done the harder work first.

I hadn't built shared clarity on who we were serving. Hadn't defined what success actually looked like. Hadn't set up signals to know if we were off course. The team was working hard on things that didn't matter enough.

So when a senior leader said we were moving slowly, I was stunned. We were shipping constantly.

But he wasn't measuring effort. He was measuring impact.

And by that measure, we were standing still.

Here's what I missed:

Getting out of the way isn't the job.

Clearing the way is.

That means building the shared meaning and context your team needs to move with conviction. The guardrails to know you're on track. Before you step back.

Empowerment without that isn't leadership. It's abandonment with good intentions.

It's the lesson I hear most from leaders I interview. We think we're going deep. We're just deep in the wrong place.

📩 More like this in Activate, my weekly newsletter — link in comments.

The relationships that stall your team…

Are the ones you’re not looking at.

Most leaders invest time with people they like or people who already support them. That's comfortable. Those relationships matter. 

But the risk isn’t in the relationships you have. It’s in the ones you haven't built. These are the people who can make or break your project. 

They are often the ones who aren’t bought in yet. 

I map relationships on two axes:
→ Do we depend on them?
→ Are they invested in helping us?

That gives me four quadrants: 

1. BUILD (High Dependency, Low Support)
 → Your blocking problem
 → Start before it’s urgent
 → Meet them where they are

2.  INVEST (High Dependency, High Support)
 → Don’t take for granted
 → Bring them in as thought partners

3. NURTURE (Low Dependency, High Support)
 → Keep warm, don’t neglect
 → Show appreciation

4. INFORM (Low Dependency, Low Support)
 → Basic updates only
 → Don’t over-invest

You're not building relationships for visibility.
You're building them so your team can move. 

 🔗 Get the 2-page worksheet (framework + mapping questions) when you subscribe. Link in comments.

I've let things go for too long more times than I want to admit. 

Not understanding what was really happening. Then coming back and realizing that I didn't know how far off things had drifted.

By then, it's really hard to bring it back.

The hardest part? Everything looked "good enough." 

Progress was happening. People were busy. Nothing seemed that broken.

"Good enough" is the most dangerous signal. 
It tells you to step back when you should be digging in.

And while you're trusting the surface, the costs are quietly accumulating:
→ Your best people leave. Not dramatically. They just see that their work stopped mattering. And they quietly find teams where it does.
→ Trust erodes. They lose confidence that you can move at the speed you need to. 
→ The window closes. While you're figuring out something is wrong, competitors aren't waiting. By the time you see it, you're already behind.
→ People burn out. Talented people don't sit idle when things aren't working. They compensate to make up for the gaps. 

The first job of a leader is to see reality clearly.

But seeing isn't enough.

You have to be willing to act on what you find. To intervene. To rebuild what's missing. To stop asking your team to compensate for what's broken.

Your job isn't to trust what looks "good enough."

It's to cut the watermelon open and do something about what you find.

🔗 I write a weekly newsletter about engineering leadership and growth. Link in bio: Melody Olson.

Empowerment isn’t just stepping back.

It’s knowing when not to.

"Step back and let them run". 

A leader I interviewed learned the hard way why this advice is incomplete.

When one of his leads went on leave, he assumed he had it covered. Smart, senior engineers. Capable tech lead. Major refactoring project already kicked off.

So he didn't check in. Didn't ask why. Didn't verify scope or timelines.

Timelines kept quietly sliding and months later, he finally looked under the hood.

The team had built a "perfect" system for a problem nobody cared about anymore.

Not a disaster. But a lot of effort that didn't move the needle at all.

His reflection: "Trust is not a substitute for verification.” 

You can trust someone's judgment and still ask why. You can empower your team and still put in guardrails to know when something is off track.

I’ve been there too and this lesson has stuck with me.

Stepping back isn't a strategy. It's the last step.

Before you let go, build the floor: 
→ Dive deep — understand where things actually are 
→ Build clarity & guardrails — so the team knows what's theirs to decide and you know if things are off track
→ Then step back

The order matters.

🔗 Get the full framework—link in comments.

I'm working on a book and people asked how to follow along. 


Starting this week, I'm sharing what I'm learning as I go.

Every Tuesday: one short email. 

What I'm discovering, building, or rethinking.

Like this week: why knowing your mission isn't the same as knowing what to build.

Link is here: https://lnkd.in/gNgUb-pg

Most companies are using a sledgehammer when they need a scalpel.


AT&T just showed what the alternative looks like.

They moved from massive general-purpose models to small, task-specific ones:
→ 90% cost reduction 
→ 70% faster processing 
→ Accuracy that met the need

Here's the thing: for most tasks, summarizing calls, parsing invoices, routing tickets, you don't need general intelligence. You need a specialized worker.

You also don't have to choose between big and small.
You can also use a reasoning model as the manager to route tasks to a fleet of fast, cheap specialists. 

This is the judgment call now:
- Where does broad reasoning matter? 
- Where is speed or lower resources the priority? 

The advantage in 2026 isn't the biggest model. It's the right architecture.

Every company has fast teams and stuck teams.

I've been interviewing leaders at companies like Netflix, NVIDIA, Meta, Datadog—this pattern keeps showing up. 

The teams that get stuck aren't full of low performers. They're full of capable people working hard, just not always on what matters most. They look good enough. And that's exactly why it is hard to spot.

What separates them? Where leaders choose to go deep:

• Seeing reality clearly (not just green dashboards) and being willing to fix it. 
• Making the stakes clear. What breaks if we don’t do this?
• Matching the right team to the right problem. Building focused teams, even when they don't match the org chart.
• Designing decision-making so teams can move quickly.

Fast teams aren’t magically different.
They’re built.

What’s one place you’ve seen leadership unlock a team?

📌 Go deep. Move fast. I write a short newsletter every Tuesday. Link in comments.

You’ll never get the right feedback unless you ask for it directly.


The feedback that changed how I lead didn’t come from a performance review. It came from a mentor I asked directly. He told me,
“You’re too focused on being nice. Be bolder. That's not helping your team.” 

At first, it stung. I cared about my team, I thought, of course I was helping them. But when I sat with it, I realized it was true. And it taught me something: The feedback you actually need usually doesn’t come unless you seek it out. 

Your manager might be thinking it. They might not know how to say it. 
Or they might be protecting the relationship instead of helping you grow. Others may give you hints, but most people won't say it directly without being asked. 

So you have to ask differently. 
→ “Is there anything I could be doing differently?” 
→ “If I were to go for promotion, what do you think the most critical feedback would be?” 
→ Ask peers, not just your manager. They see what your manager doesn’t. And you can't assume your manager will say it. 
→ Ask people who’ve moved on. They have less reason to hold back. 

The feedback that feels the hardest is usually the feedback that matters. 
Don’t wait for someone to bring it up. 
Go find it. 

That's how we grow. 

📌  Go deep. Move fast. I write a newsletter for engineers and leaders. Link in bio: Melody Olson

taplio

The Battle

Eva Karnaukh

Melody Olson

Followers

Posts

Engagements

Likes

Performance ScoreMelody Olson's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Eva Karnaukh

Melody Olson

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Melody Olson's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreMelody Olson's LinkedIn Performance