Founding Principal Engineer - Memory and Intelligence

At WYLE we are building an AI that runs the parts of your life you should not have to think about, so you get your time and attention back. Not another app. A Personal Operating System for people whose lives are too full to manage by hand.This role exists because of the hardest problem in that product. For an AI to act like you, it has to remember like you. Not a chat history. A living memory of one real person, the people in their life, their routines, their preferences, how they think and decide, and how all of it changes over years. That memory is the difference between another AI tool and a clone with conscience. It is the moat. And it is the thing you would own, end to end.What you would ownThe personal world model: a graph-native memory (we use Neo4j) that represents a single human being and everything that matters in their life as a living, evolving graph an agent can reason over.The hard problems underneath it: entity resolution, temporal state and versioning, conflict resolution when new facts contradict old ones, and decay and consolidation so the memory stays sharp over years instead of bloating or poisoning itself.The intelligence that sits on top: how the agent reasons over that memory, grounds itself in it, and learns to act more like the user over time, within guardrails.The on-device runtime: a small, fast, private intelligence layer that keeps the user's data on their device. Quantization, on-device models, low latency on a normal phone, model selection and cost at scale.This is a foundational build. The thinking behind it is real, and this layer is yours to design and own.Who you areYou have personally built knowledge-graph or graph-memory systems in production, not integrated someone else's. You think in entities, relationships, and time, not just embeddings in a vector store. Or you have genuine depth at the inference layer: small language models, quantization, on-device serving, low latency on real hardware. The best fit is strong in both, but real depth in one is what we are looking for.You go deep on one hard thing rather than wide on ten. You write the hard code yourself, and you want the keyboard. You have built inside an early-stage company before and you know what the speed, the ambiguity, and the uncertainty actually feel like, and you want that again as an owner, not an employee.How we workEngineering is horizontal on purpose. You work directly with the founder, with no layers between you and the decisions, and everyone on the team is a senior owner of something real. If you are currently a Head of Engineering, Engineering Director, Head of AI, Principal, or Staff Engineer, and what you actually miss is getting your hands back on the keyboard and owning something you build, this is for you. If what you want is another management seat, this is not.The honest partThis is a founding role and it is equity-led. You come in with meaningful ownership from the start, and we are direct about the full compensation structure early in the conversation, not after you have invested your time. If you want a salary-first role with a large team around you, this is not that, and that is completely fair.Dubai-based, or ready to relocate quickly. We provide the visa and support the move.To applyTell us about the memory or intelligence system you have built, and point us at the work. Code, a system you shipped, an open-source project, a paper with a real implementation behind it. If your depth is in temporal graphs, knowledge-graph systems, agent-memory frameworks like Graphiti, Zep, or Cognee, or a world model you built from scratch, tell us the most about that. We would rather see the work than read about it.Reach out directly at ceo@wyle.ai to start the conversation.