~

On building gnost: what 3+ years of freelancing actually looks like

March 2026 · 2 min read

I started gnost when I was 16. No clients, no reputation, no real idea what I was doing, just a Discord server where someone said they'd pay me to build an app, and I said yes before I knew how.

What I actually build

Bespoke Discord apps, built to spec. Gnost has two tracks: you can pick from pre-defined modules (economy, tickets, moderation, that kind of thing) and configure them for your server, or you can commission something fully custom for the parts that are unique to your community. Either way, you're talking directly to the person building it, not a reseller.

Most of my clients are communities that have outgrown the off-the-shelf options. Economy systems tuned to a specific game, moderation logic that doesn't exist anywhere, integrations that pull from places no existing app covers. Everything runs in Python with PostgreSQL or SQLite depending on scale. There's a one-time build cost and a monthly hosting fee. All straightforward, no surprises.

Three years in, I have a decent sense of what breaks in production. The answer is usually: whatever you were most confident about.

The thing nobody tells you

Freelance at 16 means you're doing everything. You're the engineer, the PM, the support team, and the person explaining to a client at 11pm why their leaderboard reset. You get good at scoping fast, or you burn out fast. Sometimes both.

The hardest skill I built wasn't technical. It was learning to say "that's out of scope" without losing the client.

30+ apps. 600,000+ total users across them. The revenue isn't going to make headlines, but for a 19-year-old running a solo operation from Beirut, it's been genuinely useful, and more importantly, it taught me what real software demands.

Where it's going

Gnost is still running. I take on projects when they're a good fit. But my main focus has shifted, as I want to continue pursuing my ambitions since childhood: getting deeper into systems and security. The work taught me what production software looks like from the outside. Now I want to know what it looks like from the inside.