Skip to main content
Uber logo

Schedule rides in advance

Reserve a rideReserve a ride

Schedule rides in advance

Reserve a rideReserve a ride
AI, Engineering

Transforming Financial Forecasting with Data Science and Machine Learning at Uber

July 5, 2018 / Global
Featured image for Transforming Financial Forecasting with Data Science and Machine Learning at Uber
Figure 1: Uber’s financial planning cycle occurs in three phases: strategic, operations, and insights.
Figure 2: The financial software we built at Uber incorporates multiple layers, from UI to computation to data.
Figure 3: Our financial cloud platforms enable continuous forecasting and strategic planning.
Figure 4: The directed graph is a useful means of visualizing a scenario. It shows the workflow through both metrics and computations.
Figure 5: This directed graph represents a scenario for Uber’s ride sharing business, with metrics shown in boxes and computations in ovals.
Figure 6: The Scenario Management and Computation platform, making up part of our financial forecasting system, processes scenario entities through computations to arrive at results.
Figure 7: The base scenario for São Paulo, developed with machine intelligence, combines models and metrics to determine the budget allocation required to achieve specific results.
Figure 8: The scenario accepts expert knowledge from the São Paulo city team, taking into account factors unknown by the models.
Figure 9: The financial system lets users test different hypotheses to figure out which scenario will achieve the results they want.
Figure 10: Users can override the metrics and re-compute the scenario for comparison with the base scenario.
Figure 11: Once we determine our objective and constraints, our optimization platform can build scenarios for each city team.
Figure 12: Aligning financial objectives by region lets us respond to conditions in specific markets.
Figure 13: With its objective set for maximum number of trips, the optimization platform iterates through thousands or even millions of scenarios to find the optimal scenario.
Chunyan Song

Chunyan Song

Chunyan Song is a software architect at Uber who has led the engineering initiatives to bring intelligence to several key areas: finance forecasting, fraud detection in dispatch, and incentive spend optimization in marketplace.

Posted by Chunyan Song

Category: