Presented By O'Reilly and Cloudera
Make Data Work
December 1–3, 2015 • Singapore

How Uber is using data science to make better strategic financial decisions

1:30pm–2:10pm Thursday, 12/03/2015
Data-driven Business
Location: 331 Level: Non-technical
Average rating: ****.
(4.07, 14 ratings)

Prerequisite Knowledge

Some interest in finance and passion about data science.

Description

The field of data science rose to prominence with the promise that, by using data smartly, companies could make data-driven business decisions that would outperform the decisions made in a traditional setting. There is little evidence that could prove this causality; however, there are lot of memoirs in various industries stating that data science is indeed delivering its promise.

Since its inception, data science has primarly being used in learning about products and improving engineering. Silicon Valley tech companies embed data science as part of a product org (e.g., ad clicks, ad matching etc.). Investment and corporate banks would redefine the role of quants to be data scientists and learn a lot about their products (e.g., credit card fraud, risk analysis etc). Uber also has a core team of data scientists embedded in product, engineering, and growth. As Uber is revolutionizing how people commute around the globe, it has also taken a step to bring the promise of data science into its core strategic financial decision-making.

The strategy finance team within Uber is responsible for short- and long-term forecasts, vehicle finance, and pricing. The role of pricing at Uber is to determine the optimal pricing structure that maximizes demand, while simultaneously creating a level of supplier earnings such that there is enough supply to clear the market. Accurate forecasts of supply and demand help us not only determine the optimal pricing, but also efficiently allocate financial resources. The data science team within strategy finance is primarily responsible for not only understanding the math behind all this, but also developing scalable frameworks that help us understand this across over the 300 geographies in which we operate.

The primary objective of this talk is to:

  • Share the experience of forming this data science team
  • Discuss the structure of the team, i.e. how data scientists are tightly-knitted with people from business backgrounds
  • Discuss some past projects and current problems we are working on
  • Show how the team is overcoming current obstacles and preparing for future challenges

Prakhar Mehrotra

Uber

Prakhar Mehrotra is currently leading a team of data scientists as part of the strategic finance group within Uber. Prior to joining Uber, he was a data scientist within Sales & Monetization finance at Twitter. He has an engineering degree from the California Institute of Technology.

Comments on this page are now closed.

Comments

Jordan Jordanov
02/10/2016 2:02am SGT

Hi Prakhar,
Can you share your presentation or at least parts of it (for which you think can be made public)?
Thanks