Skip to main content

Command Palette

Search for a command to run...

Become a cracked product engineer today

Qualities that make outstanding builders.

Updated
4 min read
Become a cracked product engineer today
T

Just a guy who loves to write code and watch anime.

Ship fast. Fix fast.

Speed is everything. Someone reports a bug. You fix it in 2 minutes. Not 2 hours. Not tomorrow. Right now. This is not about talent. This is about urgency. You have to feel it. The moment something breaks you should feel uncomfortable until it is fixed. It's the mindset.


Own the outcome. Not just the code.

This is the biggest shift. You stop thinking like someone who writes code and start thinking like a founder of whatever you own. You care about the bug reports. The costs. The user feedback. The whole thing. You do not wait for someone to tell you what matters. You already know because you are paying attention.

External factors do not matter. Deadlines are tight. The API is broken. The design is unclear. Vendors are hard to work with. None of that is an excuse. You find a way.

You solve hard problems because that is your job. And in 99% of cases it is doable. Especially now with the latest AI models at your fingertips. The bar for what one person can solve alone has never been higher. Own the outcome like your name is on it. Because it is.


Focus is the foundation.

Nothing works without it. You can be the most talented engineer in the room. If you spread yourself across 5 things you ship nothing meaningful. Focus means saying no constantly. It means doing one thing. Do it fast. Do it well.

This is especially true if you are a founding engineer at an early stage startup. You want to evaluate work based on effort vs ROI. Work on what's most important. Nail it. Many startups fail short because they don't understand focus.


Always ask what can go wrong.

You think you cracked a hard problem. Great. Before you commit to that solution. What are the risks. What unknowns are still there. What assumptions are you making that you have not validated. Do this before you write real code. Not after. Otherwise you're wasting time. This one habit alone saves you from shipping things that fall apart.


Get comfortable with unknowns.

You will constantly face things you have never done before. The skill is not knowing everything. The skill is moving through uncertainty without freezing. When you hit a part of the solution where you have no clue how it would work or whether it is even possible. Build something small. A quick spike. Just enough to get information about how that area of the code would look. You are not building the real thing yet. You are just killing the unknown fast so you can make real decisions.

Sometimes you don't know a thing about the topic. Then it can be worth studying the topic until you know what outcomes you want and work backwards from the desired outcomes.


Think like a user. Not like an engineer.

This is what makes you a product engineer and not just an engineer. You think about the full lifecycle. How do users find this feature. How do they use it. Where do they get confused. Where do they drop off. You do not just build the thing. You own the experience.


Data data data.

Own your numbers. Set up dashboards. Track costs. Track performance. Set up alerts so you catch problems before users do. Most engineers ignore metrics. If you do not measure it you do not own it.


Build with AI.

Prompt engineering is a real skill now. Learn it. Use AI to move faster. But this is not a one time thing. The space moves every single day. You have to be terminally online. Set up notifications on X for the best people in AI who are actually shipping and learning. Follow them closely. Read everything. The engineer who stays current with AI tooling will run circles around the one who learned it once and stopped.


Taste is a multiplier.

Many fast engineers ship ugly confusing things. Taste means you care about the details. Every interaction feels intentional. Hand crafted. Like someone actually thought about it. And honestly in the era of AI this is not hard. You have the tools to iterate and experiment fast.