Why and How to Craft Bespoke Data Science
summary
Explains how to build practical, custom data-science solutions by first measuring and preparing data (cleaning and structuring databases, defining clear metrics and benchmarks) rather than chasing vague “innovation.” Describes running rapid workshops (8-minute sketches and storyboards) to align business, developers and users, creating quick prototypes with real A/B tests, ensuring data engineers can ship production code, comparing algorithms on agreed metrics, and iterating rapidly until a deployable result is reached.
AI-written summary · about
chapters (10)
- 0:00Who has ever had a bespoke suit?
- 1:20Why make a bespoke suit?
- 9:33How to craft bespoke Data Science
- 10:41Example of an error made by Macro
- 12:58Specialize
- 16:32Don’t play with the data
- 19:10Collaboration
- 22:59Methods
- 25:17Clients
- 27:05Conclusion
related talks
25mBuilding A Successful Data Science Team
43mAgile Strategies for High-Impact Data Science & Machine Learning Projects
31mA data innovation playbook
33mManaging Stakeholders: The Key to Successful Data Science for Business
47mA Practical(ish) Introduction to Data Science
29mData Science Best Practices : From Proof of Concepts to Production
28mData Science Project Patterns that Work
43mDomain-driven Data Science