Keynote: The Culture of Data Transformation
summary
Claims that matter: concrete engineering and results. Describes that H2O rebuilt its data platform for 1k–10k+ features and redesigned storage for million‑column sparsity, that deep learning moves production to GPUs, that monolithic rule‑based systems get rewritten as microservices, that an experiment produces code via deep learning that almost compiles, that team composition reaches ~48% women, and that 40,000 fMRI studies are invalidated by a software bug, stressing reproducibility, measurable platform changes, and actual deployed tradeoffs.
AI-written summary · about
chapters (15)
- 0:00Conference opening
- 0:45Algorithms, data, business
- 3:17Software changing the world
- 5:43Building data ecosystems
- 7:29Microservices and micro businesses
- 9:17Data products and feedback
- 11:00Courage in data science
- 13:11Sparse data algorithms
- 15:16Simplifying data science
- 17:26Teamwork and high performance
- 19:34Ecosystems over products
- 21:41Open data integrity
- 24:52Storytelling and visuals
- 27:37AI disrupting software
- 29:04Closing thoughts
related talks
46mData-Driven Transformation: Organizational Changes and Leadership Insights
34mKeynote: Lighthearted Thoughts on Conversing About Data
46mPanel Discussion: Business Side of Data Science
43mKeynote: A tech görl origin story
12mKeynote: Continuous Learning: How The Fluidity of Digital Transformation...
40mMistakes were made: Data science ten years in
52mKeynote: Using Data Science to Recruit Data Scientists
40mData Engineering in a Rapidly Changing World