全职 Algorithm Engineer - Robotics Manipulation 工作, 薪水, Sharpa 公司招聘中 - Ricebowl

At Sharpa, we manufacture time by making robots useful.

Founded in 2024, Sharpa is an AI robotics company inspired by the Sherpa people of the Himalayas - known for guiding and assisting alpinists on their difficult journey to the summit. We build general purpose robots that assist, not replace, humans - freeing people from repetitive and strenuous work so they can pursue inspiring endeavors.

Our team is made up of geeks and innovators with backgrounds at leading AI and robotics companies. We are dedicated to developing ultra-high performance, autonomous robotic systems capable of adapting to the complexity of real world environments. Sharpa operates out of Singapore, Shanghai and Mountain View.

What you’ll be doing:

• Develop, monitor, and optimize learning algorithms for robots; design, train, and deploy corresponding real-time models to enable robots to perform complex physical manipulation tasks in general real-world scenarios, aiming to approach human-level dexterity as closely as possible.
• Keep abreast of industry trends and cutting-edge technological developments, and conduct pre-research and integration of new technologies and methodologies in a timely manner.

What we’re looking for:

• Master's degree or higher, preferably in Computer Science, Electronic Information, Robotics, or related fields.
• Experience in robot learning and being familiar with imitation learning or reinforcement learning.
• Experience in multimodal learning algorithms; those familiar with generative algorithms are preferred, and those familiar with large model training or usage are preferred.
• Solid foundation in robotics; familiarity with traditional robot kinematics, dynamics, and control algorithms is preferred.
• Publication in top-tier robotics conferences/journals, top-tier CV conferences, or general AI conferences is preferred; winners of prestigious academic competitions or practical experience in robotics algorithm control projects are also preferred. At least one of these criteria must be met.