Half Stack Data Science S03E05 - Code quality for data science - with Laszlo Sragner
13 snips
Oct 11, 2023 Laszlo Sragner, Director of Hypergolic, brings expertise from fintech and gaming to discuss the evolving landscape of machine learning product management. He shares insights on when machine learning is appropriate and common pitfalls like data quality and lack of engineering maturity. Laszlo underscores the importance of teaching data scientists solid programming practices and advocates for cross-functional teams to drive business alignment. He also evaluates the role of AI tools in the field, revealing that fundamental challenges in ML management remain.
AI Snips
Chapters
Books
Transcript
Episode notes
Career Path From Quant To ML Consultant
- Laszlo described a career from CS degree to quant, gaming, and then founding a startup doing NLP for finance.
- That path led him to start Hypergolic to sell machine learning product management experience.
Reframe Business Requests Before Building ML
- Reframe vague business requests into concrete ML problems before committing resources.
- Provide a roadmap and checklist that shows whether ML is the right solution or if another approach works better.
Offer Small Feasibility Projects First
- Laszlo recounts telling clients when a project is research and needs a funded investigation rather than immediate delivery.
- He offers cheap experiments to answer feasibility before committing major budgets.
