Gökhan Azap

I build products and grow them with data.

Founder of Waia, working on the business side of cloud gaming at Turkcell GAME+.

Projects

A personal finance app that tracks net worth, cash flow and investments, with an AI coach on top and a social layer around it. I designed, built, shipped and grew it alone.

App Store Google Play
- 1,000+ active users and downloads in 20+ countries, from a solo build with no team.
- Built the full product: React Native and TypeScript app on a Supabase backend with 25+ tables, row level security and 11 edge functions.
- Instrumented activation, retention and paywall funnels in PostHog, then used the drop off points to decide what to buid next.
- Ran paid acquisition on 4 channels (Meta, TikTok, Google, Apple Search Ads) against CAC and trial to paid conversion.
- Rewrote the 12 screen onboarding flow several times based on where analytics showed users dropping.
- Monetized with a hard paywall through Adapty: 7 day trial, $5.99 monthly or $29.99 yearly.
- Did ASO and a full teardown of Monarch Money, Copilot and YNAB.
- Shipped an AI coach on Gemini with voice messages and natural language portfolio management, plus live market data, Plaid bank sync and crypto wallet connections.
1,000+active users

20+countries

25+database tables

4paid channels

React NativeTypeScriptSupabase GeminiPostHogAdapty PlaidSentry
A service that turned a seller's flat product photo into store ready images of the garment on AI models. We sold it directly to sellers on Trendyol, Türkiye's largest marketplace.
- Found the customer before building anything. We ran cold outreach on WhatsApp to hundreds of Trendyol sellers, pitching better product images.
- Built a lead sourcing automation that found and qualified sellers to contact, so outreach never depended on searching by hand.
- Built the delivery as an n8n automation. One product photo in, 8 images on different AI models out, ready to upload to the listing.
- Worked with 100+ sellers, who put the output straight into their store listings.
- Built it with Hakan Göksenli. We split the outreach, the pricing, the pipeline and the support between us.

Experience

Turkcell's cloud gaming platform, powered by NVIDIA GeForce NOW. You stream PC games to a phone, a TV or a low end laptop instead of buying the hardware for them.
- Selected for Turkcell's Young Talent program and joined the business side of the product while still studying.
- Work across gaming business management, digital strategy and business development.
- Operating in a category with no settled playbook, where the unit economics are unusual because you sell access to hardware over time.
- Moving from single user analytics on my own app to cohort, partnership and distribution questions at national scale.
An early stage US startup building AI products. I worked remotely with the founder for 8 months and ran growth and marketing end to end.
- Built full acquisition funnels for several projects, from landing pages on GoHighLevel through to conversion.
- Ran paid acquisition on Meta and for Shopify e commerce clients, optimizing against CAC and ROAS instead of clicks.
- Produced the ad creatives, after going through competitors' live ads in the Meta Ad Library and adapting the angles they kept spending on.
- Designed n8n automation workflows and mapped the API integrations clients needed.
- Managed the CRM, wrote user stories and product specs, and ran requirement gathering calls with clients.
A startup inside İTÜ Teknokent. I reported directly to the CEO on the business model, the metrics that mattered and the product decisions that came out of them.
- Built the financial and operational analyses that fed strategy and prioritization.
- Went through product and process data to surface growth and efficiency opportunities.
- Prepared the investor materials and ran the DocSend data room for the fundraising round.
Two different student organizations at İTÜ, operations both times. This is where I learned that a plan nobody executes is worth nothing.

İTÜ ODER · the first registered student association in Türkiye. I joined the founding team.
- Built the promotion and operations department from scratch and ran it for 6 months.
- Designed the department end to end, from the visual identity to how events actually run.
- Ran campus field marketing, hosted promotional stands and brought in dozens of new members through direct outreach.
İTÜ Entrepreneurship Club · one of the largest student clubs at İTÜ, 2023 to 2025. Operations committee first, then the board of auditors.
- Co-led the organization of 3 major entrepreneurship events, including Girişim Zirvesi, the club's biggest one.
- Covered the whole chain on event day: sponsor meetings, hosting speakers, field coordination, stage and technical setup.

Stack

PostHog

Mixpanel

GA4

App Store Connect

Play Console

Meta Ads

Google Ads

Shopify

Sentry

Cloudflare

Supabase

React Native

Expo

Cursor

Claude

GitHub

Figma

Mobbin

Notion

About

Industrial Engineering at Istanbul Technical University. Based in Istanbul.

Everything I have done has the same shape. I start something from zero, then grow it with numbers instead of guesses. Student organizations first, then products.

I work on product and growth. I like sitting between the engineers, the data and the user, and I can build the thing myself when taht is faster than explaining it.

Outside of that I swim and cycle, with a triathlon as the goal. I write about finance and manage my own portfolio. I work in English and Turkish.

Writing

A competition run by a crypto exchange, a Bitcoin price guess and a console I did not really need. A small story about how often the answer is already sitting in public.

Bitexen, a crypto exchange, was running a competition on Twitter. Guess the price of Bitcoin on New Year's Day, closest guess wins a PlayStation 5. Hundreds of people replied with a number.

I had no view on Bitcoin. I still do not have one. What I did have was a reply section full of people who thought they did. So I collected every price they had guessed, took the average of all of them, and posted that as my answer.

It won.

It sounds cleverer than it was. Averaging a crowd is a real effect. Francis Galton wrote about it in 1906 after a country fair, where the crowd's collective guess of an ox's weight came out almost exactly right, because individual guesses are wrong in both directions and most of the error cancels out. Being the closest single answer out of hundreds still takes luck, and I had some.

I sold the console and put the money into US stocks. That became the first capital I ever managed, and it is still growing.

What stayed with me is smaller than the prize. The answer was already public. Everyone there was competing on having the better opinion. I just read waht they had already written and added it up.
What we built is not a better search engine. It is closer to a genie. You say what you want and it happens. So the genie is not the limit anymore. How much you want something is.

Look at what is in front of you. It writes the code, makes the design, pulls the data, and now it goes away for an hour and does the work before it comes back. It does not answer questions anymore. It carries out requests. Show that to someone two hundred years ago and they would not call it software. They would call it a genie.

Sam Altman said almost the same thing this summer, that we are close to building a genie that can grant any wish.

The old genie gave you three wishes, and the whole story was about not wasting them. This one gives you a hundred a day for the price of a dinner. So the hard part is not the lamp. The hard part is wanting something.

That sounds easy and it is not. Most people want things in a soft way. They would like to be richer, better at their job, finally done with the thing they keep postponing. Almost nobody wants any of it enough to sit down and ask for it a hundred times. Whoever wants more, gets more. Not becuase the genie likes them, but because they asked a hundred times while everyone else asked twice.

The second part is saying it clearly. For most of history the expensive step was doing the thing, and that cost is gone. What is left is knowing what you want and saying it well enough to be built. I ran into this with my own app. I shipped it alone and the code was never what stopped me. A genie will build the wrong product very fast.

It also never stops to ask if you are sure. It takes a confused request and hands back somethign clean and confident that looks just like a good answer. It will not save you from a bad wish.

So the gap is not access, and I do not think it is talent either. Two people have the exact same genie. One of them keeps asking for things. The other one never really wanted anything.
A monetization and retention teardown of a mobile game Soon
What Waia's paywall funnel taught me about pricing Soon
How to read the Meta Ad Library before writing a single ad Soon

Life

Let's chat

Two things I am good at: owning a growth metric on a product team, or taking a product from zero to launched. If you need eihter, email me. I answer every message.