Human Data Program Manager

About us Encord is the universal data layer for AI that helps 300+ AI teams train and run models on the right data. Our platform indexes, curates, annotates, and evaluates data across the full AI lifecycle, from development through production. Trusted by Woven by Toyota, AXA, UiPath, Zipline, and more. We're an ambitious team of 100+ working at the frontier of AI and have raised $60M in Series C funding from Wellington Management, CRV, Next47 and Y Combinator. The role We're hiring Human Data Program Managers in London to run Encord's highest-stakes human data projects end to end — the work for frontier AI labs and physical AI companies, in specialist domains such as medical imaging and robotics, and the data types we are running for the first time. These are the projects where the technical complexity is highest and where a mistake is expensive. You will work at the point where machine learning requirements become annotation work: translating what a research team actually needs into workflows, standards and instructions that a specialist workforce can execute at volume, then holding the result to a measurable standard. This is the load-bearing role in the business, and it is a hands-on one. You will know your projects' annotation standards better than anyone at the company, teach them to the annotators producing the work, decide who works on what, judge who is performing and who should come off, resolve the edge cases nobody anticipated, and answer for the delivery timeline. You will run several projects at once, and you will own how they are run. What you'll do Own delivery of your projects end to end — throughput, quality and timeline Translate complex machine learning requirements into clear annotation workflows, and design the process that produces the data the model actually needs Become the deepest expert at Encord on your project's annotation standards, and the person who resolves ambiguity and edge cases as they surface Maintain quality through process refinement, auditing and structured feedback loops, rather than through inspection at the end Train annotators onto the project and keep them improving — building the material yourself until our Learning & Development Specialist is in place Coach your annotation teams: give them the context behind the task, not only the rules, and develop their skills as the work gets harder Measure annotator performance and make the calls it implies: who continues, who needs retraining, who comes off the project Allocate tasks and manage the queue so that throughput and quality targets are met together rather than traded against each other Instrument the project at launch with the Quality Systems Lead — acceptance criteria, sampling plan, reviewer ratio — so it is measurable before the first batch ships Produce the delivery reporting for customers, and surface risk early enough that something can still be done about it Partner with Product and Engineering on process and tooling improvements, and feed