Lead Computer Vision / Perception Engineer

About OverlayOverlay is building the world’s first intelligent makeup robot—reimagining beauty application through robotics, perception, and design. As a venture-backed, pre-launch startup, we’re defining a new category of personal robotics at the intersection of self-expression and engineering. Our system operates directly on the human face, requiring an unprecedented level of precision, safety, and real-world robustness.We’re a small, highly technical team working in person in Oakland—building something that does not yet exist, and making it real.Why This Role Is DifferentThis is not a traditional execution-focused computer vision role. You will help define how perception works for a completely new class of consumer robot.This means:Identifying the hardest problems before they are obviousMaking judgment calls on what matters vs. what doesn’tDriving research, prototyping, and production—not just one layerOwning perception end-to-end in a real, physical systemYou will play a critical role in shaping both the product and the company’s technical direction.What You’ll DoWork closely with the Head of Robotics to build the the perception stack end-to-endTransition early-stage prototypes into reliable, production-ready consumer products by collaborating with external fabrication houses, contract manufacturers, and industrial design partnersDefine and prioritize perception problems that are critical to product successDesign and build real-time vision systems for precise interaction with the human faceDevelop and optimize pipelines for 3D reconstruction, tracking, and spatial understandingMake architectural decisions that scale from prototype to production hardwareIntegrate perception with robotics and motion control systems (ROS2, real-time systems)Build data pipelines, evaluation frameworks, and validation systemsCollaborate with hardware, mechanical, and design teams to deliver production systemsUnderstand and help guide the long-term perception direction and contribute to the overall product roadmapWhat We’re Looking For5+ years solving applied computer vision problems in real-world systems (robotics, embedded, or similar)Strong foundation in modern CV/ML (OpenCV, PyTorch / TensorFlow / JAX)Proficient in Python and C++ with production-quality engineering standardsDemonstrated ownership of complex systems or projects end-to-endTrack record of identifying and solving high-impact problems independentlyStrong engineering judgment — able to make practical tradeoffs under constraintsComfortable operating in ambiguity and making decisions with incomplete informationExperience translating research or prototypes into reliable production systemsClear communicator who can work effectively across disciplinesBonus PointsExperience with ROS2 and motion planning (MoveIt2, Nav2, OMPL)Background in real-time systems, embedded Linux, or hardware integrationExperience shipping robotics or hardware products to real usersHands-on prototyping experience (3D printing, machining, etc.)Strong opinions on perception systems—and the ability to defend themBenefitsHighly competitive salary & generous early-stage equityBest-in-class medical, dental & vision insurance401(k) matching programFlexible Time Off with 10 paid holidaysCommuter benefitsOffice in downtown Oakland near 19th Street BART stationThe TeamSmall, fast-moving team with high ownership and low processStrong bias toward shipping, learning, and iterating quicklyIn-person collaboration in our Oakland officeCulture & ValuesMove Fast, Move Smart: Act quickly, but with clear thinking and good judgment.Extreme Ownership: Own outcomes fully from idea to execution to results.Disagree Boldly, Then Commit Fully: Challenge ideas openly, then align and execute without hesitation.Intellectual Honesty Above All: Prioritize truth over ego. Be candid, admit mistakes, and follow the evidence.Live Demos: Show real, working results instead of talking about ideas.