DevOps for Data Science? - automate the boring stuff and leverage the OSS ecosystem
| Start time | 16:35 |
|---|---|
| End time | 17:00 |
| Countdown link | Open timer |
Have you ever developed a novel Machine learning model or application just to wait for this to be put in production - sometimes days, weeks or even months? This talk will focus on MLOps and how you can adopt these practises no matter the size of your team to enhance your idea to production workflows.
Are you working on Machine Learning or Data Science? Have you ever thought: "I am sure this could be automated. Or at least I could be optimising my workflows to make them more efficient but I do not want to do DevOps? If so this talk is for you, we will cover some MLOps tools and approaches to help you make better use of your resources while automating your Data Science workflows and their robustness.
Does your work (either research, non-profit or industry-based) depend on Machine learning, Data Science or data-intensive analyses? Have you ever wished you could automate some of the boring stuff while adding extra robustness to your workflows so that you and your team can have greater confidence and work more efficiently?
In this talk, I will present the concept of MLOps (kind of DevOps for ML scenarios, also referred to as DataOps or AIOps) and how adopting these practices can improve your team's workflows. You will learn how to automate some tasks within the ML lifecycle: from data transformation to model training, testing and validation, and deployment — making your workflows not only more seamless but your entire work more reproducible, reliable, and robust. You do not need to be a DevOps engineer to benefit from these practices, but you can indeed leverage existing open-source tools and platforms to improve your Data Science workflows. For completeness, I'll show a live end to end example, integrating MLOps practices for Machine Learning - from data processing to model training, validation and deployment. I will highlight the essential tips and tricks for each of the involved stages. You will leave the talk with practical recommendations and examples to get you started on adopting MLOps practices.
Tania is a Sr. Developer Advocate at Microsoft with vast experience in academic research and industrial environments. Her main areas of expertise are within data-intensive applications, scientific computing, and machine learning. She has conducted extensive work on the improvement of processes, reproducibility and transparency in research, data science and artificial intelligence. She is passionate about mentoring, open-source, and its community and is involved in a number of initiatives aimed to build more diverse and inclusive communities. She is also a contributor, maintainer, and developer of a number of open-source projects and the Founder of Pyladies NorthWest.