Design and build Minimum Viable Tests (MVTs) and functional prototypes to rapidly de-risk novel AI agent concepts. Execute targeted technical spikes to prove core feasibility and overcome critical architectural hurdles within strict, 4-week exploratory sprints. Apply deep expertise in agentic systems, including planning, memory, context management, tool-use, and loops, to rapidly prototype and evaluate novel solutions for design, coding and product building.
Operate with a "rolled-up mindset" by seamlessly blending the responsibilities of Software Engineering, Product Management, Research, Data Science and UX Research. Prioritize learning and prototyping over perfect code. Design clever, high-speed tests too validate User Value, Strategic Value, and Feasibility, generating raw signal before committing to heavy infrastructure. Analyze early technical and user signals to evaluate the project viability.
Job description provided by Google.
View the original posting
