A large retail bank had recently deployed a real-time interaction management platform to deliver more personalised customer experiences. While the in-house Decision Optimisation team had strong technical skills, they lacked practical experience in designing and implementing robust machine learning pipelines. Optima partnered with the team to enhance modelling expertise and embed knowledge transfer, delivering both improved performance and sustainable in-house capability.
Context:
The CCO function wanted to unlock the full value of its new interaction management platform by embedding decision optimisation and performance measurement into everyday customer journeys.
Although the Decision Optimisation team had strong technical talent, they had limited hands-on experience in building and deploying production-ready ML pipelines. This created the risk of slower progress towards a reliable Next-Best-Action framework.
Outcome:
8% sales uplift achieved through optimisation models deployed in live environments
Models improved the relevance of customer search results, boosting navigation and personalisation of experiences
The in-house team gained the capability to maintain and enhance the full data, model, and monitoring ecosystem independently
What we delivered:
Embedded Optima data science experts into the client team to co-develop decision optimisation models and NBA methodology
Designed and deployed end-to-end decisioning pipelines with a test-control framework to track sales and engagement uplift
Enhanced models to interpret customer searches more accurately, driving improved navigation and personalisation
Delivered structured knowledge transfer and upskilling to equip the in-house team for long-term ownership
Testimonial: “This project was about more than delivering stronger models. By embedding alongside the client team, we ensured measurable commercial uplift while building the operational excellence needed to sustain those improvements over time.”
— Chief Data Scientist, Optima Partners
