Research Engineer, DeepMind for Google, Energy
DeepMind's mission is to solve intelligence and use it to make the world a better place. We want to apply general-purpose learning frameworks to tackle society’s most pressing challenges, including climate change.
Progress on climate change has been frustratingly slow, in large part due to the complexity of managing massive energy systems. Where does the complexity come from? It comes from the vast, interconnected nature of the electric grid and the unpredictability of renewable energy sources. It comes from the sheer number of ways one can operate large systems, like grids, power plants and industrial facilities. It comes from the huge amount of data that these systems produce, which contain clues to radically improve efficiency - if only we could unearth them.
Traditional hand-programmed systems (i.e. rules and heuristics) struggle to deal with this complexity, which is why our work focuses on general-purpose learning frameworks: systems that learn directly from sensory experience without the need to be explicitly programmed, and which can then be applied to a wide range of real-world problems.
In 2016, we demonstrated the potential of this approach when our program, AlphaGo, became the first ever computer program to defeat a world champion at the complex game of Go, unearthing novel and creative strategies along the way. Later that year, we applied a variant of this algorithm to our first project in the energy sector - which led to significant improvements in the cooling efficiency of Google’s data centres.
Our next step is to partner with leading companies and organisations from across the industry to apply similar techniques to other large-scale energy systems.
- Work with Software Engineers and Research Scientists to translate the cutting edge research ideas at DeepMind into climate-impacting products.
- Explore complex datasets and develop machine learning-based solutions to the problems presented.
- Explore and understand the nature of energy markets and industrial systems - how they operate, what their limits are, how to interact with them safely, etc.
- Collaborate closely with a highly-distributed team. We work in London and Mountain View, and active communication is key to our successful collaboration.
- Actively engage with the team on improving the team's culture and resolving problems. None of us are ever perfect, and we will only succeed via collective self-improvement.
- At least one year working professionally with TensorFlow or other machine learning toolkits to apply machine learning to a real-world problem.
- Fluent in one or more of Google’s supported languages (C++, Java, Python, Go) and have a desire to learn more.
- Excellent communication skills. Successfully working on a distributed team requires conscious, proactive, and considerate communication in many mediums.
- Ability to work flexible hours. Our team is distributed between London and Mountain View. Most meetings take place between 3:30pm and 6pm GMT, which is 7:30am to 10am PST.
- Ability to staff production on-call rotations. We don't have SRE or dedicated oncall support, but we maintain systems that have critical uptime requirements, and we try to spread the load over all capable team members.
- Knowledge of reinforcement learning would be beneficial, but a Research Engineer comfortable with other machine learning disciplines will be able to learn this (and will have support from experienced team members in doing so).
- Knowledge of the energy industry or industrial systems would also be beneficial, but a willingness to dive into the details of complicated physical and/or regulatory systems will support rapidly learning this.
- Having experience with control systems and optimization techniques would be helpful.
We work fully in google3 and use TensorFlow as our main infrastructure. The role will suit candidates who enjoy metrics-driven work, building momentum through successive experiments and who wish to immerse themselves in some of the most cutting-edge ML and AI research.
DeepMind welcomes applications from all sections of society. We are committed to equal employment opportunity regardless of of race, religion or belief, ethnic or national origin, disability, age, citizenship, marital, domestic or civil partnership status, sexual orientation, gender identity or any other basis as protected by applicable law. If you have a disability or additional need that requires accommodation, please do not hesitate to let us know.