Sidequests
Things I built
on my own time.
I build things on the side because I'm genuinely curious and I like making things. Some of these started as experiments, some as solutions to problems I personally had. Some made money, some didn't, all of them taught me something I couldn't have learned any other way. It's also where I practice the same judgment I want to bring to backing other people's ideas: deciding what's worth building before there's any proof it'll work.
QUEST / 001
PadelGuide
Content × Affiliate × GEO Lab
I play padel twice a week and kept seeing the same problem: people struggle to find the right racket. No good resource existed. So I built one: a padel blog covering racket guides, player profiles, and technique breakdowns.
Became an affiliate partner with Padel USA, Amazon, and Padel Nuestro. Thousands of visitors, a few hundred dollars in affiliate revenue. But the more interesting thing: I can see in Google Search Console that traffic is being referred directly from ChatGPT and Claude. PadelGuide became my live lab for understanding how content gets cited by AI, which eventually led to building the GEO Analysis Tool.
QUEST / 002
GEO Analysis Tool
Generative Engine Optimisation
As AI search eats traditional SEO, most content teams have no idea how their pages perform in LLM outputs. I built a tool that scores any web page across 25 checks (citation hooks, content structure, technical access, entity clarity) and returns an AI-citation readiness score in under 60 seconds.
The framework is grounded in Princeton and IIT Delhi GEO research. Built the automation on Make.com and Airtable, integrated payments, and charged real money for it. Made a few hundred dollars. Used it on PadelGuide first. Now other people use it on their own content. It's the clearest proof I have of spotting a real shift before it was obvious, and building toward it instead of waiting for consensus.
QUEST / 003
Mahjo
Multiplayer Mahjong PWA
I learned mahjong recently and went looking for somewhere to play online. Found three or four apps, all with terrible UI, poor experience, and a $7/month paywall. The gap was obvious. So I built a better one: a PWA for multiplayer mahjong sessions, designed from scratch.
I had never written a line of code before this. Built it entirely through Claude Code and Codex, learning as I went. Now at 3,000 monthly active users, and people are already paying for it.
QUEST / 004
Everything VC
News × Directory × Founder Matching
While cold-outreaching to funds for my own move into VC, I kept hitting the same wall founders hit every day: no single place tells you who's actually deploying right now, what a fund's real thesis is beyond marketing copy, or who to actually talk to. Existing directories are either stale, US-only, or missing entirely the people trying to break into venture as a career.
So I'm building Everything VC, a news publication for the startup/VC world (founder interviews, VC interviews, funding news, startup breakdowns) with a curated global fund directory underneath it, and a career track for people trying to get into VC that nobody else in the space touches. Directory entries are hand-verified, not scraped, thesis, stage, check size, and whether a fund actually welcomes outreach.