Job Description This role is all about working as a part of our data science team analysing client data, developing new analytical features for our machine learning models and developing churn, retention, upsell and cross-sell models for new customers. Your primary responsibilities will involve: Analysing and modelling customer tabular data, predominantly using SQL, Jupyter notebooks, and Kubeflow pipelines for model deployment. Complete end-to-end model delivery process, starting from data validation and offline feature and model development, all the way to delivering production-ready feature sets and model artifacts. Building upon the standardised feature repository to assess the applicability of existing features and create new ones for our clients. Contribute to developing tools that automate data analysis, data issue identification, dataset creation, and feature selection for model training and testing. Helping our customer success team understand Telco subscriber behaviour based on interpretable propensity models to design marketing campaigns The business has developed industry-specific universal data models, which provide standardised feature sets based on data from various companies. This enables them to develop and maintain a reusable feature repository, prediction pipelines, and training those pipelines for the machine learning models. This approach allows the team to leverage existing modelling code and extend it in a reusable manner for training and evaluating machine learning models on medium to large commercial datasets. Your models will focus on predicting churn, upsell, cross-sell, and retention. Additionally, you'll contribute to the development of tools that automate data analysis, data issue identification, dataset creation, and feature selection for model training and testing.
