We built the first version of minipim while running catalog automation projects for BigCommerce merchants. The pattern was the same every time: good merchants with real products, drowning in spreadsheet logistics — import-export loops, channel rejections, localization chaos, custom field spaghetti.
The honest fix was a PIM. The problem was that the existing PIMs were designed for enterprises. Six-figure contracts. Month-long implementations. A services relationship that never ends. Our clients couldn't afford that — and they didn't need most of what it bought.
So we built the thing we would have reached for: open source, TypeScript end-to-end, an afternoon to stand up. Schema-first, headless, event-driven. Free to self-host. A managed tier for the connectors that are genuinely hard to run yourself.
The catalog is yours. No lock-in at the data layer. The PIM is a service your stack uses, not a platform your business is trapped inside.
Open by default
The engine is AGPL-3.0. Every schema decision, validation rule, and connector pattern is visible. You can fork it, inspect it, and trust it because the code is right there.
Schema is the product
A catalog without a schema is just a spreadsheet in a database. Every feature in minipim is designed around a typed, validated, channel-aware attribute model. This is the thing we got right first.
Merchants, not analysts
Channel readiness scoring exists because we got tired of watching good merchandisers get listing rejections for missing attributes they didn't know they needed. The tool should tell you what's missing, not make you discover it after a rejection.
Afternoon, not a quarter
You should be able to clone it, configure it, and have your catalog importing the same day. That's the bar we hold ourselves to. If it takes longer, something is wrong with the tool, not with you.