Jira Doesn't Work: Building an MVP in One Evening

Sunday. O’Reilly books and good music. Bliss.

For the past six years, across different projects in management roles and beyond, I’ve been stitching together a Frankenstein’s monster of tools. A patchwork of kanban, timelines, forecasts - all supposedly living in one place.

Grafana, scripts, Excel parsing, custom dashboards, report automation - and every single time, starting from scratch…

Why

The pain is simple. You want to understand what your people are actually doing right now. One tool for planning, decomposition, tracking, and forecasting.

Not three systems, not five spreadsheets, not “ask Kravtsov, he knows.”

Why not pick something off the shelf?

Because you always hit the limits. Jira is rigid and never gets updated because it’s expensive and missing the plugins you need. Half the team left and nobody’s around to maintain it. Or everything’s outsourced and you’re not allowed to change a thing.

Right now is the perfect time to consolidate everything I’ve built over the years. Contract ended, EMBA is wrapping up, thesis defense is coming. Why not.

I want one tool that shows what’s happening right now, when an epic will actually close, and what might go wrong.

One evening - a working prototype

Sat down with Claude Code, gathered all my context on delivery, roadmaps, forecasting, plus my own modules and customizations. Let’s iterate.

One evening and there’s already a working prototype. Timeline with Holt-Winters forecasting, kanban by epics, basic delivery analytics, dependencies and blockers with local model integrations via the Ollama API.

For now it’s an MVP for myself. Not a product. Not a startup. Just the tool I’ve always been missing. One that can be as flexible and extensible as I want, whenever I want.

Do you have the same pain with tooling? Or am I the only one who wants to see the entire execution on a single screen?