We're making humans invaluable in a post-AGI world.
As AI gets more capable, more of what people do every day can be done without them. We think the problem that matters is the inverse one: working out what only a person can do, and making sure those are the things people actually spend their time on.
So we're building a model of how a person works: how they think, what they value, what they'll do next. With that, the repetitive busywork gets handed off, and people are freed up for the parts that genuinely need a human. The goal is a world where the more powerful AI becomes, the more valuable each person is, not less.
We're teaching software to work the way a person does. We're currently hiring for either of the following skillsets.
Working on the capture layer: grabbing and filtering the right data through to the rest of the pipeline.
Working on the reasoning layer, turning noisy real-world capture into structured knowledge: a robust, consistent schema over irregular data, with sensible handling of ambiguity and contradictions.
A few basic questions, then we walk through a design problem together. It's more to understand your thought process and how you approach different problems, and less about your LeetCode rank.
Sound like you?
Send your CV