Tanagram is a collection of tools for developers to rapidly navigate, search, and edit code across medium, large, and ginormous repos and codebases.
The way we work with code hasn't kept up with how much we code. Companies have dozens, hundreds, or even thousands of interconnected services, sometimes in their own repositories, spanning many millions of lines of code. But our primary way of interacting with this complex system is by squinting through the few dozen lines of code that can fit on our screen at once. That's like 0.001% of the whole system (give or take a zero).
All the real knowledge — the information about how an API works, the inputs and outputs of a service, what specifically broke and paged you at 3am — exists elsewhere. If you're lucky, maybe there's documentation. If you're less lucky, hopefully you can find some time to talk with the engineer who wrote the code (if they're still around). All too often, you end up having to do your own archeology.
Tanagram maps out your whole codebase and finds all the connections. It combines call and scope graphs (powered by compiler-precise intelligence), known patterns, version control history, and custom information sources to build a complete picture of how everything works. Applications built on top of this foundation — both those that we build and those that you can build yourself — provide all the relevant context during different development activities.
Instead of only being able to see 0.001% of the system, it's like being able to see a full 1%, or even 10%, of how everything works. That's a super-power.
Tanagram is founded by Feifan Zhou, who was an early engineer on Stripe Issuing (and experienced this codebase context problem firsthand as Stripe's codebase grew to about 10 million lines of code). Tanagram is backed by Pear VC (PearX W25).