We design, ship, and maintain the distributed systems that power real-time AI inference, petabyte-scale search, and verified knowledge graphs across 140+ languages.
Performance, reliability, and developer velocity drive our technology choices. We standardize on battle-tested infrastructure while remaining pragmatic about trade-offs.
Every article, citation, and multimedia asset passes through a multi-stage verification and indexing pipeline before reaching users.
Our technical decisions are guided by principles that prioritize correctness, transparency, and long-term maintainability.
We ship fast, but never at the cost of data integrity or user trust. Every feature includes automated regression tests and canary deployments.
Internal tools, RFCs, and postmortems are public unless they contain PII or security-sensitive logic. Transparency builds better systems.
We use AI to automate boilerplate and surface insights, but human experts retain final authority on knowledge verification and editorial policy.
Multilingual support isn't a feature—it's a constraint. We design schemas, UIs, and search algorithms from day one for 140+ locales.
We contribute back to the ecosystem and run our development process openly. Every major change starts with an RFC.
Every incident is a system failure, not a person failure. We document root causes and implement preventive controls.
Decisions are documented, not discussed in meetings. PRDs, RFCs, and runbooks replace status updates.
Engineers dedicate one day per week to tooling, research, or open-source contributions aligned with platform goals.
Team spans 14 time zones. We optimize for written communication, overlapping cores, and async code reviews.
We're looking for engineers who care about correctness, love building infrastructure, and want to work on systems that impact millions of learners worldwide.