I'm considering spending the next one or two months building one substantial open-source project instead of several smaller portfolio pieces. The idea is an AI-assisted developer productivity platform for planning learning goals, organizing daily tasks, tracking progress, and integrating with GitHub. In later versions, I might add repository analysis, architecture diagrams, code explanations, and debugging suggestions. Does this sound like a worthwhile project, or is it trying to solve too many unrelated problems at once? I'd appreciate honest feedback on how to narrow the scope or make the project more technically impressive and genuinely useful.
4 Answers
The biggest issue is that this is currently a feature list rather than a clearly defined problem. Most of these capabilities already exist in editors and development tools, so you need to explain what specific frustration your project solves better than the alternatives. Pick one audience and one painful workflow, then build a focused solution around that.
I’d avoid starting with an all-in-one productivity platform. Choose one narrow use case, such as connecting a reported bug to relevant logs, commits, and code locations, then producing a useful first-pass diagnosis. That gives you a concrete input, measurable output, and a realistic way to demonstrate technical depth.
If your goal is learning, building the larger idea can still be worthwhile, but define a small first release. Spend the first couple of months on one complete workflow, such as GitHub issue triage or turning a learning goal into trackable coding tasks. You can always expand after proving that the initial feature works and that people actually want it.
Repository analysis and architecture diagrams could be interesting, but only if the output is actionable rather than just a generated summary. For example, the tool could identify risky dependencies, explain why a component is difficult to change, or highlight areas affected by a proposed modification. The value should come from helping developers make a decision, not simply adding an AI label.
That makes sense. I’ll try to choose one workflow with a clear user problem instead of attempting to build the whole platform immediately.