Founding Applied Research Engineer | AI & Knowledge Systems

What if you could work on the part of AI that sits underneath the application layer?Join a stealth, early-stage AI company building foundational technology for how organisations structure, retrieve and reason over complex information.They're tackling a problem that becomes increasingly important as organisations adopt AI: having access to more information isn't enough. The real challenge is understanding relationships, preserving context and being able to reason reliably across large volumes of messy, changing data.We are backed by strong technical talent and is building a small, highly capable team around this problem. They're now looking for an Applied Research Engineer to take ownership of the systems that sit at the core of the platform.This is an opportunity to work somewhere you can take research ideas, turn them into production systems and have a direct influence on the technical direction of the company.What you'll get to work onYou'll design and build retrieval systems that go beyond basic semantic search, combining dense, sparse and structured approachesYou'll explore how information can be represented so that AI systems can reason over entities, relationships and contextYou'll build information extraction systems capable of dealing with ambiguous, messy and unstructured dataYou'll work on entity recognition, relationship extraction, coreference resolution and temporal understandingYou'll develop approaches for maintaining and evolving knowledge as new information becomes availableYou'll work on provenance, validation and conflict resolution so information can be trustedYou'll stay close to relevant research and translate useful ideas into production technologyYou'll have significant influence over the architecture and technical direction of these systemsWe're looking for someone whoHas research or industry experience in information retrieval, NLP, knowledge representation, knowledge engineering or a closely related fieldHas taken research beyond experimentation and into a real-world systemThinks deeply about how information should be structured to support reasoningUnderstands that retrieval is about more than simply finding the most similar documentsIs comfortable working from first principles and defining problems without a predefined roadmapEnjoys working in a small team where you'll have significant ownershipIs pragmatic and cares about shipping useful systems rather than research for research's sakeYou could come from academia, an applied research team or a highly technical engineering environment. The key thing is the depth of your thinking and your ability to turn that thinking into something that works.Why join?Work on foundational AI problems: Get beyond the application layer and work on how AI systems actually understand and reason over informationResearch → production: See your ideas move from papers and experiments into systems used in the real worldHigh ownership: You'll have genuine responsibility for a core part of the technologySmall technical team: Work closely with experienced founders and engineers without layers of managementTechnical influence: Your decisions will directly shape the architecture and direction of the platformIntellectual challenge: Work on difficult problems around retrieval, knowledge representation and reasoning that don't have obvious answersEquity: Have meaningful ownership in an early-stage companyThe environmentYou'll be joining a small, founder-led and deeply technical team where you'll be expected to think independently, challenge assumptions and move quickly.There isn't a huge amount of process or bureaucracy. You'll have a high degree of autonomy, but you'll also have a high bar for technical thinking and execution.The team works hybrid from London, with flexibility around how and where you work.The company is currently operating in stealth, so further information about the product and customers will be shared during the interview process.If you're interested in NLP, retrieval, knowledge representation or reasoning and want to work on a genuinely foundational AI problem rather than another wrapper around an existing model, I'd be happy to tell you more.