The problem of a modern data delivery platform: you want employees within your organization create value from data by developing models for investments, forecasts etc. But these employees do not have IT skills, other than knowing a language like R or Python, so the models they create live their own lives, copies exist and soon nobody knows which model to trust. How do you bring together the two worlds of citizen development and stable IT? How can you implement an open source software like workflow for models, with pull requests on repositories holding the model code, automated builds, tests and deployments? How can you package generic algorithms into shared libraries for reuse?

In this session we will take you through lessons learned at a financial institution where we dealt with exactly this problem. We will show you how we applied Databricks, Synapse, Data Factory and other Azure services to implement a modern data platform where citizen development is brought together with IT.