Industrial Data Scientist

Recycleye is a London-based technology company that develops AI-powered sorting equipment and analytics solutions for the waste management industry. Its core products include robotic sorting systems and data-driven insights tools designed to help recycling facilities improve throughput volume, increase offtake purity, reduce operational expenditure, and address labour shortages on sorting lines. The company serves waste management businesses globally, offering solutions capable of identifying and sorting a wide range of materials including plastics (PET, PP, HDPE), non-ferrous metals such as aluminium, fibre, cartons, and waste electrical and electronic equipment (WEEE). Recycleye's mission is rooted in transforming the world's waste into a resource, contributing directly to the circular economy by enabling more accurate and efficient material recovery. By retrofitting existing recycling infrastructure with advanced computer vision and robotics, the company helps divert greater volumes of valuable materials from residual waste streams, supporting sustainability goals for both its clients and the broader materials sector. What distinguishes Recycleye is its combination of retrofittable, robust AI robotic systems with real-time analytics, making advanced automation accessible to existing plants without requiring full facility overhauls. Its platform supports limitless waste classification categories and has been deployed in partnership with leading industry operators, demonstrating measurable improvements in sorting accuracy and plant profitability.

The opportunity
Recycleye and CPG are building a team to provide data analytics and actionable insights to a waste facility operator. There are near Infrared sorters (NIR), balers, AI powered airjets, AI robots, mechanical screens, and many different types of machines in a waste facility. The waste that comes in is varied on a daily, hourly, minute basis. The machines and plant can be configured in thousands of ways to sort the material, changing conveyor speeds, sorter settings, and coming up with complex business logic to match output to material pricing and revenue.

This is an exciting opportunity to join the team to help push the boundaries of material sorting in waste facilities - understand the data, present to a team of experts, and develop models to automate and optimise throughput, revenue, and material output purity. If you ever wanted to make a difference in the world of recycling this is it!

Responsibilities Overview

• Track, measure, analyse billions of rows of time-series data from a waste facility. Motors, sorters, AI cameras, etc.
• Develop methods to link, simplify, clean/de-noise the data and visualise in clear graphics/analytics for a non-technical audience (experts first, waste operators second) to understand and action.
• Develop plant-level understanding - from the data. Translate data into logical rules or first principles for improved understanding.
• Build statistical models detect anomalies and optimise plant output, revenue, etc. and test them. Understand the tradeoffs.
• A/B Test different plant setups based on optimisation algorithms, measure output and iterate.
• Develop scoring functions, feedback loops, and quantitative metrics to compare results against.

Requirements

• A background in data analytics and statistical modelling from data.
• Ability to write high-quality production-level code (we use Python), mathematical models, statistical simulations. Advanced SQL skills.
• Comfortable with getting hands-on in large scale physical systems. Going into waste facilities to deeply understand processes, flow of material, the gap between physical and virtual world.
• Experienced in data visualisation
• Practical - ability to get into the customer shoes and solve trivial breakdown or other problems before complex optimisation.
• Ability to travel (to the US) up 1-2 weeks a quarter.

It's a bonus if you have

• Experience in Reinforcement Learning or similar optimisation techniques
• Machine learning experience
• Computer vision experience - especially object detection models
• Experienced with large datasets - billions of rows.
• Time-series data experience
• Anomaly detection