Avatar for Optimal

Deploying autonomous indoor farms outside every city on earth

Model Based Reinforcement Learning Engineer

£60k – £110k • 0.0% – 1.0%
Apply now

We are a collection of engineers and scientists from Oxford, MIT and DeepMind, on a mission to grow safer, healthier food by deploying fully autonomous indoor farms outside every city on earth.

Our team includes a professor of control engineering, a research scientist who helped reduce the cooling bill of a Google data centre by 40%, and a farmer who started as a vegetable picker 40 years ago and now runs one of the most advanced indoor farming operations in the world.

We are backed by world-leading deep technology VC funds, including Founders Fund, who have backed companies such as SpaceX, Palantir and Square from the very start. We are well capitalised for the future having raised one of the largest seed rounds ever in Europe.


What conditions do you give a plant in the next time step to maximise the long-term nutritional quality of its produce? How do you achieve those conditions most accurately whilst minimising the use of scarce resources such as fresh water and energy? These are some of the important questions you will help answer, working in a multidisciplinary team across machine learning and reinforcement learning.

Horticulture is an art based on human intuition. You will be turning this scarce intuition into scalable science - creating superhuman crop-growing algorithms and deploying them on a massive scale, to provide the world access to safer, healthier food.


- Model, simulate, explore and optimally control biological and physical systems, exploiting our deep, proprietary dataset.
- Understand and model the uncertainties within our environment and devise strategies to reduce them.
- Safely transfer algorithms from simulation environments to our large-scale, $50m indoor farms.
- Work with our professor-level research supervisors to analyse results and iterate approaches.
- Work closely with our machine learning and software engineers.
- Continually learn and self-improve, helping others to do the same.


- PhD, advanced degree (or equivalent experience) in engineering, computer science, physics or a related field.
- Strong foundations in probability, statistics, linear algebra, signal processing, optimisation and control.
- Experience with bayesian and frequentist approaches for time series modelling: GPs, RNNs, Bayesian NNs, etc.
- Strong data manipulation, exploration and visualisation skills.
- Strong programming skills in python and relevant libraries: numpy, pandas, matplotlib, tensorflow, pytorch etc.

Extra credit
- Experience deploying model-based RL algorithms into real-world environments.
- Extracurricular activities: contributing to the open-source community, side projects, coding competitions etc.

More jobs at Optimal

View all jobs

Machine Learning Research Engineer

Apply now

Forward Deployed Software Engineer

Apply now