Troubleshooting Agile DataSciOps?
Jul 14, 2021
The podcast delves into the challenges faced by isolated data science teams and highlights the importance of close collaboration between engineers and data scientists. It explores the evolution of operational practices, the integration of data science in product development, and the significance of teamwork for successful project delivery. The episode also discusses the potential for collaboration between developers and data scientists, emphasizing the benefits of pairing individuals for enhanced productivity and efficiency.
AI Snips
Chapters
Transcript
Episode notes
Data Science Teams Are Isolated
- Data science teams are often isolated like early 2010s system admin teams, creating a disconnect from the main development pipeline.
- This isolation causes challenges in product integration and understanding, similar to early dev and ops silos.
User Clustering Reveals Personas
- Jeffrey shared an example where a data scientist clustered users by usage patterns to derive personas for product improvement.
- This collaboration led to validated hypotheses and insightful new usage patterns.
Disconnected Efforts Cause Failures
- Lack of collaboration leads to unusable data science outputs and frustration as developers struggle to adopt delivered models.
- Data scientists' efforts can be wasted when their work is disconnected from production realities and workflows.