Sharing the Story of a Failed Project (Developer)
Last year, I made an AI chatbot as a side project. The initial plan was good, but...
1. I used an overly complex tech stack. Included MSA, Kubernetes, and LangChain. It only increased management complexity.
2. Without any user research, I just kept adding features, so it didn't match user needs.
3. I got exhausted developing alone and gave up halfway.
Lessons learned: launch quickly with a small set of features, get feedback, and lower your tech ambitions. Now, based on my failure, I'm rebuilding it little by little.
Everyone, take note when working on side projects!
8 answers
Actually, MSA only makes sense at scale; doing it alone can be counterproductive. I also suffered when I first introduced Kubernetes. Now I'm keeping it simple.
Isn't the biggest issue that they didn't do user research? Just adding features is a dangerous mindset. But since they're reflecting on it and rebuilding, I'll cheer them on.
Agreed, I've had a similar experience.
This is right lol. If you get too greedy with features, you'll ruin it.
Do you have a source?
Well... the idea of building an AI chatbot solo with MSA is a bit much; but thanks for sharing what you learned.
I've also failed a few times, but starting small is really important. Right now, I'm working on building an MVP first.
You worked hard, ㅠㅠ