Machine Learning Engineer – Operational Research

The Role

As a Machine Learning Engineer, you will play a crucial role in the development, implementation and maintenance of cutting-edge machine learning products. Your responsibilities will involve engineering sophisticated machine learning models, as well as refining and updating existing systems.

In this team, you will develop the algorithmic and machine-learning systems that power Deliveroo’s delivery network. You will work in a cross-functional team alongside engineers, data scientists and product managers to develop systems that make automated decisions at massive scale. The team has independence and works on some of the most interesting problems at the intersection of riders, consumers, and restaurants. We evaluate the performance of all our decision-making machines through our world-class experimentation platform.

You will:

• Optimise our delivery network by making rider assignment decisions; predicting how long a leg of the delivery journey will take; or mitigating real-time delays.
• Enhance our simulation capabilities to more accurately predict the effects of algorithmic changes on our delivery network.
• Optimise consumer and rider fees.

Also, you will work alongside people who work on:

• The consumer experience by showing the most relevant restaurants and dishes.
• Detecting fraud and abuse from consumers, riders, and restaurants.
• Assisting restaurants in optimising their presence on Deliveroo, for example by recommending that they improve their menus or photography, or add a popular dish.
• Creating an ML platform to improve our ML and optimisation capabilities.

You will report into a ML/OR Manager. This is a hybrid role that will be based inLondon.

Requirements:

• You are someone who knows the fundamentals of machine learning and operational research and when they should be applied through a relevant PhD or work experience.
• You can translate fuzzy logistics and delivery problems or objectives into a well-thought-out algorithmic solution. You get satisfaction from seeing your algorithms shipped and driving measurable impact to the business.
• Experience in programming, where the work involves programming with Python, Rust and Go.
• Experience in discrete event simulations and/or combinatorial optimisation problems.
• Understand end-to-end model productionisation.
• A bias to simplicity, where you care most about achieving impact.

Nice to haves:

• Experience in solving Vehicle Routing Problems (VRP) and/or building large scale delivery network simulations
• Experience in any of the following areas: algorithms and data structures, parallel and distributed computing, high-performance computing