Scientific Product Manager

Within EIT, the AI & Robotics Institute (AIR) is looking for a Scientific Product Manager, i.e., a product manager able to lead at the interface between innovative science, frontier AI and state of the art product thinking. You will be collaborating with the scientists who use our systems, the software, robotics, design and AI engineers who build them, and the AIR and program leadership to decide where development goes next.

Your role will be to decide which problems are worth solving, find the relevant solutions that solve them, and anticipate how to push these solutions further.

You will work with subject matter experts, developers and AI researchers. While your role is not to replace them, you do need to be scientifically fluent enough that scientists trust your read of their work and technically fluent enough that engineers trust your read of solution design and implementation.

• →Work with scientists to identify and prioritize scientific problems to maximize impact on the research cycle (time from question to result, information gained per experiment, reliability of the outcome)
• →Act as the “voice of the users” for your team and leadership, to surface needs at three levels: those users state, those they experience but do not articulate and those likely to emerge as throughput, automation or model capability shifts.
• →Translate needs into problem statements they and the engineers can act on, covering success criteria, constraints from the biology and the hardware, available data, acceptable failure modes, and what a human must sign off. Conversely translate the solution’s capability in limitations into trade-off the users can understand and weigh to make informed decisions.
• →Advocate for the minimal viable solution addressing the need and quick iteration. Not everything needs an AI model, but it is important to identify when it’s the case. Avoid over-engineering and low-quality AI-generated solutions. Identify opportunities from latest developments in AI, robotics, lab automation.
• →Help identify the data a solution needs: what exists, its quality and provenance, and the gaps that require new experiments.
• →Identify product derivation opportunities (tools, services or datasets with value beyond the originating program) and build the early case for them, working with the relevant teams on IP, data rights and commercial considerations.

• A background in experimental science, ideally life sciences (including, but not limited to, molecular & cellular biology, microbiology, plant biology), with first-hand knowledge of how lab work actually runs and where it goes wrong.
• Working technical fluency across AI/ML, software and automation. You know which classes of method suit which problems (rules, regression, deep learning, LLM agents) and what data each needs. You can read a model evaluation and spot overfitting, leakage, or a benchmark that does not reflect the bench. While you do not need to build models, ability to autonomously (including AI assisted coding) develop quick prototypes to test hypothesis is a plus.
• A history of taking a technical solution from problem framing through to adoption by real users, in an R&D, product or consulting setting.
• Structured problem solving and strong quantitative reasoning, including experimental design and working statistics.
• Effective communication with academic scientists, engineers and executives alike, including the ability to disagree constructively with each and steer towards consensus
• Comfort with ambiguity and ownership in a growing R&D organization, where requirements are discovered through building.

• Experience as a translator, product manager, product owner or technical consultant on AI or data products, ideally scientific ones.
• Exposure to laboratory automation, robotics, LIMS or ELN systems.
• Experience turning bespoke projects into platform capabilities or products, or exposure to technology transfer, IP or commercialization.
• Familiarity with user research methods such as contextual inquiry and workflow or journey mapping.

• Resourceful and persistent through technical, organizational and political roadblocks
• Motivated by how much your team’s solutions improve scientists’ work, with a strong sense of ownership, a bias towards action and short build cycles.
• Curious about both science & technology with a desire to keep learning.
• Skeptical of complexity, deriving satisfaction from removing parts rather than adding them.
• Collaborative enabler, working as a peer to engineers and designers, not a gatekeeper between them and users.