Built with:
Just the back-end of it here. SplitPay let's clients create bills that need to be split amongst friends. You create a bill, add people to it, have each person indicate what they've paid for and what they've consumed, and let SplitPay generate an optimized set of transactions.
Unfortunately, the problem of optimzing transactions is NP-Complete. I'm not going to go into that here, but you can show a reduction from either the Partition problem or Subset-Sum problem to prove it.
As such, I used a greedy approximation algorithm instead. Not sure what the relative performance guarantee is, or if it's even a ρ-approximation. Alternatively, one can model this problem as a max-flow problem and solve it with Dinic's or Edmonds-Karp but I just want to keep things simple.
I'll write this soon enough once I decide how I want to deploy and dockerize the app.
Kinda lazy to write unit and integration tests, so I went with e2e tests for the highest ROI. I have a Postman test suite that I use to test locally and also on the CI pipeline to test each build on Master. You can check out a live version of the test suite here:
