Week 3: When a Paid Project Derailed My Indie Dreams

Build in Public — Week 3 (May 4–10)

Right after the architecture reset, an opportunity came up that I couldn’t turn down. A former contact needed help with a medical document AI project and asked if I could take it on. It was paid, it was interesting, and after months of spending on domain names and API credits without any revenue, the timing felt right.

I said yes.

The project itself was fine — straightforward requirements, reasonable timeline, and the client was someone I had worked with before. I spent about five days on it, delivered what they needed, and got paid. It felt good to know that the skills I had been building — working with AI models, structuring prompts, building small prototypes — had real market value.

But those five days came at a cost. They were five days I didn’t spend on Lumi. And in the early stage of a solo project, five days is a lot. A feature that could have been finished stayed in the backlog. A bug that could have been caught early got discovered later. The momentum I had built from the architecture reset slowed to a crawl.

The Juggling Problem

The hardest part wasn’t the work itself. It was the context switching. Every time I opened Android Studio, I had to remind myself where I left off in the Lumi codebase. Every time I switched back to the consulting project, I had to reload the requirements and reorient myself. Split focus meant both projects moved slower than they should have.

I tried different approaches: work on the consulting project in the morning and Lumi in the afternoon. That just meant I was bad at both. I tried dedicating full days to one thing — that worked better, but it meant Lumi got pushed to the end of the week, and by Friday I was too tired to do quality work on my own project.

What finally worked was batching: dedicate blocks of consecutive days to a single project, no mixing. Three days on consulting, four days on Lumi. It wasn’t perfect — I still lost some momentum — but it was better than switching every few hours.

The Honest Moment

I needed the money. That’s the simple truth. Building a product takes more than time — it takes domain registrations, API credits, testing devices, and a dozen other small expenses that add up. A consulting project that covers those costs isn’t a distraction. It’s what makes the indie work sustainable.

But I also missed building my own thing. There’s a different kind of motivation when you’re working on something you own. The features matter to you personally. The design decisions are yours alone. Every line of code is an investment in your own vision, not someone else’s. Those five days reminded me why I started Lumi in the first place — not because I wanted to build software for clients, but because I wanted to build something that people would use in their daily lives.

What I Learned

Week 3 taught me that solo development isn’t just about technical skills. It’s about managing energy, focus, and financial reality at the same time. Taking paid work isn’t a failure of commitment — it’s part of the journey. The key is to batch it, contain it, and never pretend you can do everything at once.