Brandon Brown

As engineering organizations scale, so does the cost of data inconsistency. Teams multiply, languages diverge, and the subtle misunderstandings between producers and consumers accumulate until something breaks in production. At SecurityScorecard, we confronted this head-on by adopting a schema-first engineering culture — and building the open source tooling to make it stick. This talk is a practitioner's guide to building a centralized schema codebase that works across multiple programming languages and database technologies. We'll cover how we use buf cli and bash scripting to generate cross-language client libraries from a single source of truth, what schema compatibility checks actually look like in a CI/CD pipeline, and how we integrated with DataHub to give both frontend and backend teams a shared view of the contract landscape — so frontend engineers can build against generated types the day a proto lands, without waiting on backend. We'll dig into the real decisions: how to structure a schema monorepo, how to enforce compatibility without slowing teams down, and how ownership lines get drawn when a schema change touches five services. We'll also cover what this looks like…