Abe Gong, Data Scientist, Jawbone
Data science borrows tools from many related fields: statistics and machine learning, "big data," software development, etc. Together, they open new opportunities for finding insight, building data products, and creating value. But the practical work of bringing together diverse tools from multiple disciplines presents new challenges as well. Many organizations---even those with talented teams and good infrastructure---still struggle to effectively unlock the power of data.
Borrowing from experiences at cutting-edge data companies including Jawbone, LinkedIn, and Concurrent, this talk shares strategies and examples for aligning data science workflows to create more value in less time. The Sidekick Pattern---using small, carefully curated data sets to multiply the power of big data---gets special attention as a versatile and often-overlooked technique for streamlining data workflows.
Understanding data science workflows is the key to making good use of day-to-day development time. It is also a prerequisite for making smart investments about hiring, building infrastructure, and prioritizing projects.
If you are a data scientist, this talk will help you accelerate your workflow and build more high-value things faster.
If you are a specialist in a related field, this talk will help you understand what's new about data science, so that you can borrow tips, tricks, and tools for your own practice.
If you are a manager of data scientists, this talk will help you to train, support, evaluate, defend, align, and inspire your team.