How I work.
In practice.

The process

1. 01

   Get into the room

   Before any brief or wireframe, I try to be where the problem actually lives. With Savior, that meant months inside hospitals before touching a spec. The gap between how things are supposed to work and how they actually work is usually where the product is. You can't see that on a call.

2. 02

   Find the actual problem

   The brief is usually a symptom. "We need a dashboard." Okay, but what decision is it serving? I spend time here before writing anything down. The real problem is almost never the stated one.

3. 03

   Build the smallest useful thing

   Not an MVP in the buzzword sense. The smallest thing that actually does the job for one real person. Career Leap started as a WhatsApp group. KnowYourPay started as a spreadsheet I shared with a few people. If the thing doesn't work before it's a product, it won't work after.

4. 04

   Ship early, watch what happens

   I have a bias toward shipping early over debating internally. Real users in 3 days beats a perfectly refined prototype in 3 weeks. The feedback you get from watching someone actually use something is worth more than any survey or interview.

5. 05

   Know when to kill it

   I shut down Career Leap when the unit economics didn't work. I killed Hypersync before it launched when the third person I showed it to said "I just use Doppler." The willingness to stop is as important as the willingness to start. Sunk cost is the enemy of good judgment.

How I think

• —

  Doing beats planning. A working thing teaches you more than a perfect document. I'll always bias toward making something real, even if it's rough.

• —

  Small teams are a feature. Career Leap peaked at 3 people. Savior was tiny until it wasn't. Overhead kills speed, and speed is usually the advantage.

• —

  Distribution is part of the product. How people find it, share it, and explain it to someone else. That's not a marketing problem. That's a product problem.

• —

  Honest numbers over comfortable ones. I'd rather know that retention is bad at week 2 than celebrate week 1 activation. The uncomfortable metric is always the useful one.

• —

  Shutting down is a decision, not a failure. Career Leap didn't work at scale. That's data. Acting on it is what kept me from wasting another two years on something that wasn't going to work.

What I don't do

• ✕ Long discovery before any output. A week of honest conversations is worth more than a six-week framework.
• ✕ Pixel-perfect prototypes before validation. Teams fall in love with Figma files. Real users interact with real things differently.
• ✕ Vanity metrics as north stars. Downloads and signups only matter in relation to something real. Who came back, and why?
• ✕ Consensus-driven decisions. Feedback is an input, not a vote. Someone has to make the call. That's the job.