Supply-chain planning
Better plans from the planning engine you already have.
For companies that build planning software: forecasting, inventory, production planning.
Your product already turns data into plans. The rules, heuristics and pre-processing steps inside it decide how good those plans are, and how fast they come back. That is the code we improve.
- Forecasting pipelines and the models inside them
- Inventory and replenishment rules
- Production-planning heuristics, and the steps that shrink a model before a solver runs
- Forecast error (for example WMAPE or RMSE)
- Plan cost or service level
- Time to produce a plan
- Kinaxis An AI-improved heuristic that shrinks planning models scored about 2× better than hand tuning, on Kinaxis’s own measure. Tested on one problem so far. Source
- Kinaxis A forecasting pipeline got 22.6% lower error and took 90% less time to run. Source
- Coolblue 28-day demand forecasts got more than 5% more accurate, after about 200 tries. Source
These are other companies using Google’s AlphaEvolve. They show that this kind of code has room to improve. Argmax adds the part they had to do themselves: proving each gain is real.
The checks are the same everywhere
Old and new code run on the same problems, several times each. A statistical test rules out luck, nothing else may get worse, and some problems stay out of the AI’s reach so wins must hold up there too. Every accepted change arrives as a pull request your engineers review.
Other use cases
Machine & job-shop scheduling
For companies that build scheduling software for factories: jobs on machines.
See the use case → Available nowPacking & cutting
For companies whose software packs or cuts: bins, containers, pallets, sheets, rolls.
See the use case → ExploringBeyond optimisation
For teams whose code has a clear speed or size number: compilers, databases, ML training.
See the use case →