We figured out that small companies would try it but they didn't need the solution. Large companies needed it desperately but wouldn't share data with an untested startup.

Later, when I was building my own process mining tool, I hit the same Catch-22.

This demo was my attempt to side-step the data concerns and avoid enterprise IT onboarding altogether.

I built it in Python and used Pyodide to run it entirely in the browser. Pyodide provides the CPython interpreter compiled to WebAssembly, so the Python code executes client-side.

Getting pandas, numpy, and visualization libraries working in WASM was tricky. Initial load takes ~15 seconds, but after that performance is good. The constraint visualization updates in real-time as you add process steps.