OpenHack DevOps for Data Science (OHDTSC)
Who should attend
Experts with different skill sets and workflows across data engineering, data science, machine learning engineering, development, operations, and other knowledge areas are working together in a collaborative way emphasizing the skills and strengths of each team member.
- Software Engineers
- Cloud Solution Architects
- Data Scientists
- ML Engineers
Knowledge Prerequisites To be successful and get the most out of this OpenHack, familiarize yourself with the following: Data Science:
- What is Machine Learning? (website)
- What is Azure Machine Learning? (website with video)
- Basic familiarity with Jupyter notebooks (website with tutorial)
- What is DevOps? (website)
- Continuous Integration
- Continuous Delivery
- DevOps Practices
- Azure DevOps Overview (website)
- Comfortable with git basics or git hands on learning (only Main: Intro sequence, Ramping up, Remote: push & Pull sections)
- Understand Azure Pipelines basics or Create your first Azure Pipeline Do YAML, not classic (website + code sample)
- Azure DevOps Branch Policies (website)
- • Participants should have familiarity with programming languages like Python.
Tooling Prerequisites To avoid any delays with downloading or installing tooling, you are encouraged to have the following ready to go!
This OpenHack enables attendees to employ fundamental up to advanced DevOps practices for the Data Science process, leveraging Azure Machine Learning Service, Azure DevOps, Azure Data Factory, and other relevant Azure services. This OpenHack simulates a real-world scenario where an insurance company needs to predict the probability that a driver will initiate an auto insurance claim in the next year and needs to be able to take the Data Scientist’s local functional model and associated data used to train the model to production in a high-quality, secure, scalable way.
During the “hacking”, attendees will focus on:
- 1. Understanding DevOps fundamentals as applied to the Data Science process to train and deploy machine learning models
- 2. Begin to apply more advanced DevOps practices (such as canary rollout or taking automated actions based on instrumentation)
By the end of the OpenHack, attendees will have built out a technical solution that automatically trains, evaluates, registers, and deploys a model, connects fundamental DevOps practices for the Data Ops for ML used to train the model, as well as implements observability aspects for the system.
This open hack provides a real-world context within which the various data science, machine learning, and DevOps capabilities within Azure can be applied to an end-to-end data science scenario; enabling participants to compare and validate options.
- Entry point for people who don’t know ML
- Upskill for “DevOps” with ML Devs & Data Scientists