Case Study: MandM Direct achieves 10x more production models and scalable, traceable MLOps with Dataiku

MandM Lightning Fast Value From Data Science With Dataiku

MandM Direct, one of the UK’s largest online retailers with over 3.5 million active customers, struggled with siloed data and the rising complexity of deploying and maintaining machine learning models (initially run as .py files on local machines and GCP VMs). To scale analytics and make ML deployments transparent and traceable, MandM Direct partnered with Dataiku and Google Cloud Platform (BigQuery) to democratize data access and streamline model operations.

Using Dataiku on GCP, MandM Direct built visual data pipelines, MLOps workflows, and push-down compute to BigQuery, enabling roughly 10x more models versus a code-only approach and running hundreds of live models with clear performance visibility. Dataiku helped create a 400+ feature library, broaden access to nontechnical users, separate design and production environments, and save “literally days of work every month,” driving measurable impact across pricing, personalized marketing and customer lifetime value initiatives.

MandM Direct

Ben Powis

Head of Data Science