ML Engineer

ML Engineer (Stealth AI Company) About the company We are building a foundational intelligence platform that transforms fragmented, proprietary information into durable institutional intelligence - enabling organisations to reason faster, preserve context, and compound knowledge over time. We are starting with information‑dense, judgment‑heavy industries where decision‑making under uncertainty is core. Long‑term, the platform is designed for any information‑led organisation where trust, provenance, and context matter. Our focus is not surface‑level AI features, but the intelligence substrate that workflows depend on. The problem we're solving Most organisations don't struggle with data volume. They struggle with: fragmented information across systems and time loss of context and institutional memory repeated manual synthesis knowledge walking out the door AI tools that retrieve information but don't reason over itWe are building the foundational layer beneath workflows: how information is structured, contextualised, and reasoned over. What we build We build software that helps organisations understand their own information, not just store or search it. The platform: ingests internal and external data structures information to preserve meaning, relationships, and provenance enables reasoning across time, sources, and uncertainty keeps humans in the loop where judgment matters evolves as organisational knowledge evolvesWe are intentionally not: a workflow automation tool a chat UI on top of documents a standalone "knowledge graph product"Graphs, ML, probabilistic reasoning, and human‑in‑the‑loop systems are combined to solve a larger problem: How can organisations reason reliably over their own information at scale? The role As an ML Engineer, you'll work at the intersection of machine learning systems, knowledge representation, and reasoning infrastructure - helping build the core intelligence layer of the platform. This is not a model‑tuning or API‑wrapping role. You'll tackle foundational problems such as: Knowledge extraction & structuring Designing ML pipelines that turn unstructured, proprietary data into semantically rich representations. Reasoning systems Building and integrating models that support probabilistic reasoning, multi‑hop inference, and context‑aware decision support. Agentic workflows Developing systems where AI augments human judgment via explainability, uncertainty estimation, and feedback loops. Evaluation & reliability Defining metrics and testing frameworks appropriate for high‑stakes, information‑led environments. Production integration Working closely with backend engineers, product, and domain experts to ensure ML systems are robust and scalable.What you'll be expected to do Design, train, and deploy ML models that handle real‑world complexity: noise, ambiguity, evolving schemas Think deeply about information representation, not just retrieval or ranking Contribute to architectural decisions around ML infrastructure and system design Ship working systems, iterate based on feedback, and avoid over‑engineering Maintain a high bar for clarity, reproducibility, and long‑term maintainabilityWhat we're looking for Strong foundations in machine learning (e.g. NLP, information extraction, representation learning) Systems‑oriented mindset - performance in production matters more than benchmarks Comfort working in ambiguity and defining problems from first principles Intellectual honesty and willingness to challenge assumptions Motivation to build infrastructure that compounds in value over timeNice to have Experience with graph databases (preferably Neo4j) Background in information retrieval (search, ranking, semantic search, hybrid systems) Experience building or operating ML systems in enterprise cloud environments, particularly AzureWorking environment Based in London In‑office by default with work from home on Wednesdays Founder‑led, deeply technical, and substance‑driven Low‑ego, high‑ownership culture Strong opinions, fast feedback loops, and a high bar for clarityMinimal ceremony, maximum focus on building durable systems. Values First‑principles thinking - design from fundamentals Human judgment matters - AI supports decisions, it doesn't replace responsibility Intellectual honesty - correctness over hype Trust by default - security, provenance, and explainability built in Compounding advantage - systems that get better over time Build foundations, not wrappers - infrastructure over surface features Services advertised by Gold Group are those of an Agency and/or an Employment Business. We will contact you within the next 14 days if you are selected for interview. For a copy of our privacy policy please visit our website