About
A practical intelligence layer for distributors.
Built to solve a specific operational problem, not to be the most sophisticated piece of software in the supply chain stack.
Why this exists
The decision problem in distribution.
Ethan Byrne · Founder & CEO
Allodial Supply Co started with a straightforward goal: supply products to local businesses efficiently. The intention was to understand distribution from the inside, covering sourcing, pricing, logistics, and customer behavior, before building software around it.
Through that work, a different problem became visible. Distributors were not losing customers in a fight. They were losing them quietly — an account that used to order every three weeks slowed to five, then eight, and nobody said anything because nothing looked broken on any given day.
By the time a name is missing from a quarterly report, the competitor has had a full cycle to settle in and the call is a much harder one. Allodial Predict exists to make that call weeks earlier, while there is still something to save.
What guides this
Company
Which accounts are flagged, at what level, in what order, and the sentence explaining each one are computed arithmetically inside PostgreSQL. No model participates in any of it, so the same data always produces the same list. Separately and deliberately stated: the in-app assistant (Allobot) and the import column matcher do send operational data to OpenAI under its API terms. Neither decides anything about an account, and none of your data is used to train a model.
Want to run a Replay on your data?
Drop in an order export and the same rules run over your history, in your browser. Free, and there is no account to create.