Abstract: Machine learning has become an integral part of modern business operations, but the success of these projects depends on the quality of the underlying software. Unfortunately, many machine-learning prototypes fail to reach production systems because data science teams incur accidental and intentional technical debt faster than they get to their solution.
We have largely solved the problem of under-engineering our machine-learning prototypes by leveraging software engineering best practices. This talk will showcase open-source tooling like Kedro to create production-ready machine-learning code.
Attendees will leave with a clear understanding of how software engineering can empower data science teams to deliver better results and drive business value by increasing collaboration, reducing errors and improving the efficiency of their code.
The future is maintainable and modular machine-learning code.
Attendees should have baseline knowledge of Python and do not need to be familiar with Kedro
Bio: Yetunde Dada is the Director of Product Management at QuantumBlack, an AI-focused branch of McKinsey. She is instrumental in building products for Data Engineers and Data Scientists, including a notable Python library known as Kedro. Kedro is a distinguished product, marking the first open-source offering from McKinsey and QuantumBlack.
She holds an MBA from the Said Business School at the University of Oxford, earned in the 2017/2018 academic year. Her professional background is diverse and includes roles such as Data Engineer and Data Product Manager at Absa (formerly known as Barclays Africa Group Limited), Innovation Consultant at Engineers Without Borders South Africa, and a Mechanical Engineer.