Staff Machine Learning Scientist (Recommendations)
Team: Engineering & Data
Location: Depop - London
Company Description
Life is about creating. That's why we're home to over 30 million artists, stylists, designers, sneakerheads — and you? We're the community-powered, circular-minded marketplace changing the world of online fashion. Now it's time to get inspired at Depop.
Responsibilities
The Recommendations team builds models that power discovery at Depop, helping millions of users find items that they will love. As a Staff Machine Learning Scientist, you’ll set the technical vision for our next-generation recommendation models, lead high-impact initiatives, and mentor others to drive innovation at scale.
Responsibilities
You will:
Lead the design and deployment of advanced recommendation systems, encompassing encoder-based architectures, vector representations and large-scale retrieval.
Mentor, coach, and set technical direction within the Recommendations team, helping others grow and innovate.
Collaborate closely with cross-functional partners (product, engineering, data) to define problems, translate them into scalable solutions, and deliver measurable business outcomes.
Lead the end-to-end lifecycle of ML projects: from ideation, data acquisition, feature engineering, training, and evaluation to deployment and ongoing iteration.
Drive innovation in recommendation systems by researching and integrating emerging ML techniques, frameworks, and tooling, while contributing technical expertise to long-term product and data strategy.
Act as a thought leader in the recommendations space, sharing learnings internally, engaging with the wider ML community, and showcasing our work externally.
Qualifications
Proven track record in designing, deploying, and optimizing large-scale recommendation systems, including candidate retrieval and ranking models, with measurable impact in production environments.
Deep understanding of machine learning fundamentals and applied experience with architectures including collaborative filtering, deep learning, and hybrid recommendation approaches.
Proven ability to productionize ML models and pipelines: from prototyping to de
ployment, with strong experience in monitoring, iteration, and troubleshooting.
Advanced programming skills in Python and familiarity with ML frameworks such as PyTorch, TensorFlow, or similar.
Solid foundation in stats, experimental design, and working with offline/online evaluations in real-world settings.
Experience leading projects and mentoring engineers or scientists, with a track record of fostering team growth and technical excellence.
Excellent communication skills: able to bridge technical and non-technical stakeholders and influence decision making.
Committed to responsible AI practices, including attention to ethics, fairness, and inclusivity.
How we work
MyMode is our new hybrid-working model, designed to empower our employees to choose a working mode that works for them.
MyMode is composed of 3 working modes: Flex, Office Based and Remote.
Flex (Default)
Tell me more +Show me less -Flex is our default working mode, meaning all employees will automatically enrol in this mode and there is no application required to enter this mode. Flex employees will be expected to work from the office at least 4 days per month. Teams will determine whether there are set weekly or monthly in-office days based on their operating rhythms and practices. You will need to work with your manager to determine your in-office schedule for your team.
Office Based
Tell me more +Show me less -This option is for employees who are committing to work from the office for a minimum of 4 days per week. As part of taking on the Office Based working mode you will be able to apply for a permanent desk in the office if you need one, but you won’t need to apply to become an Office Based employee.
Remote
Tell me more +Show me less -Under the Remote working mode you are able to work anywhere within the country you are employed in. This mode requires around 2- 4 days per year in the office, depending on organisational guidance. You will be able to expense travel if you are asked to attend the office, but not for office attendance by your choice.
*Remote working is not applicable for all roles at Depop, please check with our Talent Team.
Application Process
Our DNA encompasses the central reasons that people are proud to work at Depop and unites us with a shared language and sense of community.
It guides our daily interactions and empowers individuals, teams, departments and our company as a whole to have a greater impact and achieve our mission.
Show up for the community
Tell me more +Show me less -We go above and beyond. When they succeed, we succeed.
We’re changing how millions of people buy, sell and explore their style, so we do everything we can to create a safe space in a community where you can learn, grow and succeed on your own terms.
Have each other's backs
Tell me more +Show me less -We empower each other with kindness and respect our differences.
Everyone at Depop is seen, heard, valued and encouraged. Our genius is born from our diversity of thought, so we celebrate our wins together and hold each other up when things get tough.
Act with purpose
Tell me more +Show me less -We take conscious risks, deliver efficiently and learn from our mistakes.
Our mission is to be the world’s most diverse and progressive home of fashion. We have the conviction to succeed, the patience to learn and the confidence to fail and try again - being open all the way.
Think thrift
Tell me more +Show me less -We’re resourceful, seek out opportunities and we hustle.
We’re powering a future that is more thoughtful, circular and better for people and planet. To do it, we stay curious, savvy, resourceful and empowered to get the job done – effectively and responsibly.
At the heart of our mission...
At the heart of our mission...