3 Session Tracks
Data Science Management
Best practices and transformational strategies to make data science leaders and their teams more productive and satisfied, while increasing knowledge sharing and efficiency.
- Understanding and removing the hidden barriers to enterprise data science
- Navigating the journey to a model-driven business
- Managing your team across the enterprise data science life cycle
- Standardizing processes across data science teams & tools
- Practical strategies to prioritize and apply AI projects
- Increasing the capacity of your data science teams
- Compounding data science value with user adoption & collaboration
- IT & data science partnership strategies for success
Applications of data science, machine learning and MLOps supporting organizational strategy, operational efficiency and significant business value.
- Don’t fight fires, prevent fires: Using ML to change the paradigm
- Leading from the top: The role of the CEO in driving data science
- Where you should be exploring next with data science?
- Model Velocity, what it means, and why you need it
- Moving the needle on business adoption and allowing for failure
- Data scientists as storytellers: Making the math exciting
- When IT and data science work together at scale
- Industry panel on ethics, transparency, AutoML, and tracking
MLOps Tips, Tools and Techniques
Techniques, tools, and advice across the entire MLOps lifecycle from exploration and experimentation to deployment and monitoring.
- Practical updates and advice on popular tools
- New research on modeling approaches
- AI ethics & explainable AI for deep nets?
- Practical updates and advice on NLP and transformer-based models
- Tips for making transfer learning work in the wild
- Finding the right acceleration framework for the job (Spark, Ray, Dask, RAPIDS, etc.)
- How to get the most out of GPU-based analytics
- Best practices for working with data science IDEs
- The latest applicable trends in model monitoring
- Tips for establishing viable model validation flows
- How to get the most out of source control repositories and CI/CD pipelines