Robotics Engineer（Physical AI）

◾️About APTO APTO is a startup established to solve the “data preparation challenges” faced by companies developing AI. We provide a cloud-based data collection and annotation platform that streamlines the creation and management of high-quality datasets required for AI model development and operations. Our core services include: - An AI data platform that centrally manages data creation by both internal and external workers - Training dataset provisioning across a wide range of domains - Contract-based solutions supporting AI model development These services are increasingly being adopted by major enterprises. In addition, through our smartphone-based crowdworker network and multilingual, multi-regional data collection infrastructure, we are building a system capable of supporting global AI deployment. ■Background of the Position APTO is currently accelerating investment into the Physical AI and robotics domain. As a result, not only AI model development, but also the ability to design hardware systems capable of acquiring, controlling, and operating real-world data has become increasingly important. In this role, you will work closely with AI and software teams to drive the development of hardware systems supporting Physical AI, including: - Mechanical design using CAD - Prototyping - Sensor/device integration - Enclosure design - Real-machine testing - Simulation This is not simply a contract-based design role. You will be involved in a 0→1 phase where we define and build the hardware required for the AI era from the ground up. ◾️Responsibilities - Mechanical and enclosure design using CAD - Prototype development, assembly, and evaluation - Hardware architecture design including sensors and cameras - Design and improvement of robotics-related equipment - Device development for AI data acquisition - Specification coordination with software and AI teams - Component selection, procurement, and vendor coordination - Design improvements with mass production in mind - Hardware improvements from PoC to production operations - Data augmentation ◾️ Why This Role - Opportunity to join the launch phase of the “Physical AI” domain, one of the next major areas following generative AI - Work closely with AI/ML engineers across both software and hardware domains Involved end-to-end from prototyping and assembly to improvement and operations — not limited to CAD design - Participate in projects focused on actual AI data acquisition and robotics deployment, not just PoCs - Significant ownership over specification design, technology selection, and hardware architecture in a 0→1 environment

- Experience in 3D CAD design using tools such as Fusion360, SolidWorks, CATIA, or Inventor - Practical experience in mechanical, enclosure, or jig design - Experience handling the full cycle from prototyping and assembly to evaluation and improvement - Experience designing with consideration for 3D printing, machining, and sheet metal fabrication - Experience working with hardware configurations including sensors, cameras, and electronic components - Hands-on experience improving actual hardware rather than only desk-based design - Experience collaborating with software engineers and electrical engineers

- Experience developing robotics-related products - Robotics development experience using ROS / ROS2 - Experience developing sensing devices using cameras, LiDAR, IMU, etc. - Experience developing data acquisition devices for Computer Vision or AI - Prototyping experience using Arduino, Raspberry Pi, or Jetson - Experience in control and validation using Python or C++ - Experience in mass-production design, manufacturing outsourcing, or ODM/OEM management - Experience building products from 0→1 in startups or new business environments - Strong interest and expertise in Physical AI, Robotics, or Edge AI

- Someone who excels at structuring ambiguous customer requirements into clearly defined problems - Someone interested in roles spanning business, technology, and project execution - Someone who can act proactively without being constrained by role boundaries