Senior Robotics Perception Engineer

• Max Ahrens (Founder & CEO): ML PhD + Postdoc (Oxford University); led AI research projects with the Alan Turing Institute and UK Ministry of Defence; ex-McKinsey.
• Simon White (Founding Software Engineer): former YC startup founder/CTO; ex-defence systems engineer.
• Joel Kang (Founding Product Engineer): former AI startup founder/CTO; ex-LinkedIn; product-minded engineer focused on shipping end-to-end systems.

Our Culture

1. Nerdy and Intellectually Athletic: We dive deep into first principles and embrace heterodox thinking to solve the hardest technical problems.
2. Strong Opinions, Loosely Held: We debate with data and logic, then commit fully.
3. Continuous Improvement: Direct feedback, low ego, high trust, high growth.
4. Outcome Obsession: Results over activity. You own your path and deliver high-impact results.
5. Innovation Through Diversity of Thought: We seek diverse perspectives and backgrounds to build innovative, world-class solutions.

Your Role

• As Maihem's Senior Robotic Perception Engineer, you will architect the vision and metrology foundations that make adaptable robotics possible. You'll design, build, and integrate the perception stack into our downstream planning, learning, and control systems.
• You'll transform raw sensor data into sub-millimeter accurate geometry for collision-free robot planning and navigation. You'll apply your expertise in 3D computer vision, sensor fusion, and deep learning to plan, build, and deploy our perception stack.
• This is a hands-on role: you'll design the sensing setup, implement the core 3D pipeline, and ensure it's robust enough for production deployment with our industry partners.

What you’ll do

• Sensor evaluation and integration: Experiment with different hardware setups, and assess how their mounting requirements and constraints affect the quality of the data captured.
• Calibration and frame correctness: intrinsics/extrinsics, hand–eye, frame conventions, repeatability tests; build tooling to validate, detect and correct drift.
• Image registration: Align multiple perceptual data sources (e.g. RGBD, point-clouds, TSDFs, geometries) to a CAD model using appropriate registration and outlier handling techniques (feature extraction, RANSAC, ICP, robust kernels etc).
• Digital twin update + ML pipeline inputs: Ensure CV/perception data feeds accurately to digital win simulation and machine learning pipeline work streams.
• Field robustness: logging/replay, health checks, and failure mode detection for deployment in industry partner cells.

Must-haves

• 3+ years shipping real-world CV/3D systems beyond prototypes (robotics, inspection, metrology).
• Hands-on experience with depth sensing (stereo/RGB-D/ToF/structured light/laser profiling) and the real-world tradeoffs.
• Strong geometry fundamentals, including:
  • coordinate frames and transforms, rotations/quaternions
  • geometric error handling, including reprojection and calibration errors
  • registration concepts (feature-based + ICP-style alignment), outlier handling
• Built calibration + validation tooling (repeatability tests, drift checks, confidence metrics)—not just “run the library.”
• Strong fundamentals of programming in a team.
• Comfortable debugging messy reality: reflective metal, smoke/spatter, lighting variation, partial occlusions.
• Proven team player, clear communicator; you think first principles, you identify assumptions and test them methodically, you question the ‘why’ behind every decision.

Nice-to-haves

• Experience in harsh industrial settings (welding/fabrication/heavy industry).
• Robotics integration exposure (robot frames/tooling, latency constraints, ROS2/MoveIt-style patterns).
• Edge deployment (NVIDIA GPUs, TensorRT/CUDA), latency and reliability monitoring.
• Synthetic data or simulation integration for perception evaluation (e.g. Isaac Sim).
• Familiarity with or interest in reinforcement learning pipelines (e.g Isaac Lab).

Benefits

• Private health insurance + pension contributions
• Free lunch + gym
• Curiosity allowance (books, learning, memberships)

A note on our in-office culture: Our team works in the office five days a week—and we love it. It's a core part of our energetic startup culture. Please only apply if that excites you too.

A meeting lands on your calendar

If they’re a yes, I book the chat. You show up — that’s the whole job-hunt.