Lead ML Engineer — Production AI & LLMs

Requirements

• Experience deploying multiple machine learning models into production

,

• 5+ years of experience in data science, machine learning engineering, or related roles

,

• Experience integrating and evaluating LLMs

,

• Excellent knowledge of both Data Science (Python, SQL) and production tools

,

• Understanding of probability and statistics fundamentals

,

• Strong ability to communicate findings to non-technical stakeholders

,

• Experience of leading projects involving multiple people including developing a short term roadmap and reporting progress

,

• Comfortable breaking down work incrementally

,

• (Desirable) Familiarity with Docker containers and container orchestration tools

,

• (Desirable) Experience with LLM-as-a-judge and/or annotation pipelines

,

• We strongly encourage applications from people of colour, the LGBTQ+ community, people with disabilities, neurodivergent people, parents, carers, and people from lower socio-economic backgrounds

What the job involves

• Machine Learning Engineers at Cleo work on building novel solutions to real-world problems

,

• This really does vary but could be: creating chatbots to coach our users around their financial health, creating classifiers to better understand transaction data or even optimising transactions within our payments platform

,

• They train, deploy, and improve machine learning models in production, ensuring they deliver meaningful impact for our users

,

• We’re looking for our next Lead Machine Learning Engineer to join our Chat Evaluations team

,

• You’ll shape & guide technical work within a team of adaptable, creative and product-focused engineers, who deliver ML/AI features that improve the observability of chatbot quality and power the AI development cycle

,

• Here are some examples, big and small, of the kinds of product feature work our ML Engineers have taken part in over the last year:

,

• Deployed best-in-class credit decisioning models which affect millions of customers, using open banking data rather than traditional credit scoring

,

• Developed models to interpret transactional data, enhancing the understanding of users’ finances. Think about your bank statement—how often do you not recognise a transaction on first review?

,

• Fed user interaction data into fine-tuned LLMs and contextual ranking models, so Cleo knows how to start and continue an engaging conversation

,

• Developed optimisation models to improve payment success rates for customers while minimising business costs, tackling this as a two-sided optimisation challenge

,

• Designed and implemented AI agents to analyse and extract insights from users’ transactional data

#J-18808-Ljbffr…