Founding Head of Data& Engineering

Some data problems support a product. Others define what the product is.

An AI tutor that actually teaches needs more than a strong language model. It needs to know what to teach, in what order, through which approach, and whether the student truly understood. That is not just a prompt engineering challenge. It is a knowledge and data modeling challenge — and one of the hardest in education.

We are looking for a Founding Head of Data& Engineering to define the core knowledge and data layer of an AI-native learning product from the ground up.

There is no inherited infrastructure, no existing schemas, and no legacy system to work around. What you build will become the foundation everything else depends on: the knowledge structures, the content systems, the instrumentation layer, the experimentation framework, and ultimately the way the system understands how humans learn.

Design and build a knowledge graph that represents not just what to teach, but how to teach it: which concepts unlock others, where learners typically get stuck, which learning paths work for which kinds of students, and how to encode all of that into a structure a machine can navigate across subjects, languages, and scale.

Own the end-to-end data architecture for a new kind of product: knowledge representation, content systems, content generation workflows, student interaction capture, and real-time learning session data — all built from scratch, all designed to operate at very large scale.

Create the measurement and experimentation layer that helps the organization understand whether learning is actually working: how to compare teaching approaches, how to measure understanding rather than surface completion, and how to distinguish meaningful outcomes from noise across millions of interactions.

Work closely with pedagogy, product, and AI teams to translate human teaching expertise and learning theory into production-grade models, schemas, and systems.

Strong experience building or transforming complex data infrastructure, including making foundational architectural decisions and owning their long-term consequences.

Experience modeling domains where the structure was not obvious — where understanding the domain was the hard part, not just implementing the technology. Background in education, healthcare, legal, or other messy human-centered domains is a strong advantage.

A mindset that treats data not as a pipeline, but as a representation of a real-world system. You care about the domain itself. You want to sit with pedagogy experts, debate what matters, and make thoughtful decisions about what the system should represent.

Broad hands-on experience across the data stack, from modeling and architecture to instrumentation, experimentation, analytics, and enabling infrastructure.

• Leadership experience is a plus, but this role is especially suited for a player-coach — someone who can shape direction, mentor others, and still loves building deeply as part of a high-talent founding team.