Software Engineer, Google DeepMind
Robotics Simulation
AI systems to solve problems and advance scientific discovery
Job no longer available
AI systems to solve problems and advance scientific discovery
1001+ employees
Job no longer available
1001+ employees
Company mission
To build AI responsibly to benefit humanity.
Role
Who you are
- Graduate level theoretical background of numerical computation, mechanics and geometry
- Numerical computation: Convergence and stability properties of numerical optimization routines. Good understanding of floating-point performance considerations on current CPUs
- Mechanics: Basic Lagrangian mechanics, conservation principles
- You must be very comfortable with C and C++
- Some familiarity with articulated-body simulators: e.g. Bullet, MuJoCo, PhysX, Vortex, ODE, Havok
Desirable
- Further background relevant to rigid body simulations:
- Mechanics: Kinematics using quaternion algebra, Theory of contact and friction, Linear Complementarity Problems
- Geometry: Convexity, collision detection, ray casting, intersecting volumes, convex decomposition
- Numerical computation: Integration of Ordinary Differential Equations
- Experience with writing physics simulation code yourself (perhaps in an experimental, academic capacity), and/or familiarity with multiple articulated-body simulators
- Experience with porting numerical algorithms to GPU
- Participation in open-source projects
What the job involves
- Develop advanced physics simulation infrastructure in C, C++ and Python, including numerical computation, research-friendly APIs, performance optimisation, rendering and visualisation
- Collaborate with a tightly integrated team to advance the capabilities of the software stack, starting as a team member and taking on increasing leadership responsibilities over 12 months
- Collaborate with multiple research groups across DeepMind, Google, and externally, including several robotics teams
- Track and manage an ongoing database of bugs, issues and feature requests, with a requirement for careful, thoughtful prioritisation
- Manage open-source projects with multiple contributors
Salary benchmarks
Share this job
Insights
Company
Funding (1 round)
Jan 2012
$5m
ANGEL
Total funding: $5m
Our take
Google DeepMind combines the talents of two leading AI labs, Google Brain and DeepMind. Over the past decade, they have made major breakthroughs in artificial intelligence, which have helped shape the AI industry today.
DeepMind first gained attention with its deep reinforcement learning techniques, especially with the program DQN, which learned to play 49 Atari games just by watching the screen. Its landmark achievement came with AlphaGo, the first AI to beat a world champion in the complex game of Go. Following this success, it tackled even tougher challenges, such as creating AlphaStar, which defeated professional players in StarCraft II, and AlphaFold, which predicts protein structures to advance medical research.
Now operating as Google DeepMind, the organisation uses its computing resources to solve important scientific and engineering problems. Focused on developing AI responsibly, ensuring that its groundbreaking products benefit everyone, With ongoing innovations in areas like generative models and robotics, Google DeepMind is set to lead the next wave of AI advancements while addressing the ethical challenges that come with these technologies.
Steph
Company Specialist at Welcome to the Jungle