Decision making: Uncertainty and automation

Talk abstracts

Decisions automation at scale

How can we leverage quantitative models and their implementation to make decisions in real time and at scale in an online food delivery marketplace? There are three key actors involved in food delivery: diners (ordering food), restaurants (preparing food) and delivery providers (transporting food from restaurants to diners). In orchestrating the actions of these actors, there are a number of key challenges involved: demand prediction, contracting the right number of delivery providers, coordinating handoff of food at restaurants, accurately communicating timing estimates to all the actors and smart routing that can account for multiple business needs.

Efficient food delivery systems require automation of these tactical and operational decisions at scale. This is achieved through effective use of data and quantitative models to power the systems that make those decisions. We always focus on empowering execution driven systems making consistent and repeatable decisions. During this talk, attendees will hear about some of the challenges in this space and how we manage this at Grubhub.

What are the odds? Decision making under uncertainty

We live in a complex world where many business decisions have to be made under uncertainty i.e. a decision maker has to choose an action based on often imperfect observations and potentially with little understood outcomes. Automated decisions, using AI, have to be made under uncertainty too and it all depends on how AI perceives and handles uncertainty. In this talk, we will look at the problem of decision making under uncertainty, discuss the possible approaches to the problem and how to communicate uncertainty to decision makers.

Speaker bios

Sagar currently leads a team of operation researchers, data scientists and developers at Grubhub. His responsibilities include creating models to automate decision making at scale in the food delivery marketplace and building realistic simulation capabilities. Specifically he oversees routing algorithms, scheduling and runtime predictive models. Sagar has a breadth of experience in developing machine learning models and designing practical optimization algorithms in multiple industries. Prior to Grubhub he spent 5 years in the options trading industry building statistical signals for optimal automated trading strategies and portfolio optimization. Sagar holds a PhD in Physics from Northwestern University. His research there focussed on dynamical evolution of complex networks. He devised local systematic perturbation strategies that appeared to be harmful at local level but would eventually drive the network to a desirable global state. Prior to that Sagar completed his undergraduate education in India from Indian Institute of Technology, Madras.

About Ensemble

Ensemble is an artificial intelligence, machine learning and data science event held regularly at Peak in Manchester. Like an ensemble model, the event will allow learning through exposure to a diverse set of experiences.

Each month, a top researcher from the fields of computer science, statistics or operational research will be invited to present their work. The audience will be made up of data scientists, software engineers, researchers and company CIOs from tech companies and universities.