Data Scientist, Machine Learning

  • Monzo
  • London, UK
  • Dec 19, 2018
Full time Data

Job Description

We're looking for a curious, adaptable Data Scientist to join the data team at Monzo who is excited to help build the best bank account in the world. You'll have the opportunity to analyse, build, deploy and improve the machine learning systems that delight our customers and help our customer support agents to operate swiftly and efficiently.

About Monzo

At Monzo, we're building a bank that is fair, transparent and a delight to use. We’re growing extremely fast and have over a million customers in the UK, with over 60,000 new people joining every month. We’ve built a product that people love and more than 80% of our growth comes from word of mouth and referrals.

We are looking for people who are passionate about changing the way the industry works and who share our values. We believe in an open, transparent, highly diverse working environment where everyone is empowered to make changes. Following Stripe’s example, all emails in the company are available for everyone to explore, and we discuss everything in Slack channels which are accessible by everyone in the company. You can see more about our plans on our public product roadmap, and sneak peeks of upcoming features in our community forum.

The data team

The data team's mission is to

Enable Monzo to Make Better Decisions, Faster

This mission encompasses three major areas of work: (1) product analytics, to help teams understand our customers and improve our app (2) domain analytics, to support teams who are working in specific banking disciplines (e.g., lending, finance, and financial crime), and (3) machine learning, where we design and build system that automate decisions across Monzo. While we take a flexible approach and frequently help each other across these areas, we each have one domain that is our primary focus.

For this role, we are looking for a Data Scientist who will focus on machine learning. The machine learning sub-team partners with teams across all of Monzo to design, build, and experiment with machine learning systems that help them achieve their goals. For example, see our recent work with related articles in the app help screen, and the approach we took when building the help screen's search algorithm. This sub-team is currently two Data Scientists and one backend Engineer.

The role

You’ll be the third Data Scientist to focus on machine learning, working with many teams across Monzo to help them to design, build, analyse, and experiment with machine learning systems that make use of the data we gather.

We organise our machine learning projects into three distinct phases; you'll spend your time doing all three:

  • Explore. We use BigQuery and Jupyter Notebooks to analyse data and design machine learning models for offline evaluation. For example, we are investigating machine learning powered assistants for our customer operations team, classifiers to detect financial fraud, and NLP models to improve the app's help screen.
  • Launch. We build Python micro services and cron jobs to put promising machine learning models into production. We are actively working on automating as much of this step as possible: our goal is for any Data Scientist to be able to deploy a promising new model to production in less than a day.
  • Iterate. We run A/B tests in partnership with other teams and analyse the results. Based on outcomes, we may decide to roll the model out to every customer or to explore improvements to the model for further testing.

What’s special about data & machine learning at Monzo?

Autonomy. We believe that people reach their full potential when you can remove all the operational obstacles out of their way and let them run with their ideas. This comes together with a strong sense of ownership for your projects. At Monzo, you will get full access to our data and analytics infrastructure. When you discover something interesting, there is nothing stopping you from exploring and implementing your coolest ideas.

Cutting-edge managed infrastructure. All our data infrastructure lives on the Google Cloud Platform, so you don't need to spend your time configuring or managing clusters, databases, etc. If you want to train a Machine Learning model faster, just spin up a compute engine instances and submit a job from your local machine, no DevOps skills required.

Automation. We aim to automate as much as we can, so that every person in the team can focus on the things that humans do best. As with all data science work, there’s some analysis and reporting, and as much as possible we encourage self-serve access to our data through Looker.

You should apply if:

  • What we’re doing sounds exciting, and you can’t wait to explore our data
  • You're impact driven and eager to have a real positive impact on the company, product, users and very importantly your colleagues as well
  • You have a self-starter mindset; you proactively identify issues and opportunities and tackle them without being told to do so
  • You're a team player whom your colleagues can rely on
  • You have a solid grounding in SQL and Python, and are comfortable using them every day
  • You’re happiest exploring data, designing and evaluating machine learning models and seeing these projects all the way through to production
  • You're excited about the potential of machine learning and can communicate those ideas to colleagues who are not familiar with the domain
  • You’re adaptable, curious and enjoy learning new technologies and ideas

Logistics

We can help you relocate to London and we can sponsor visas.

We care deeply about inclusive working practices and diverse teams. If you’d prefer to work part-time or as a job-share, we’ll facilitate this wherever we can - whether to help you meet other commitments or to help you strike a great work-life balance.

Our interview process typically consists of a 30 minute initial phone screen, a take-home test, and a half-day on-site interview. We promise not to ask you any brain teasers or trick questions, and we won't make you code on a whiteboard