arrow_back Product development for fleet management
Predicting Order Batchability
Submitted by sunil rathee (@ratheesunil) on Saturday, 20 October 2018
Section: Crisp talk Technical level: Intermediate
For a business looking to provide delightful user experience in the online food delivery space, it is paramount important to meet the promises. This talk will be mainly focusing on the time compliant deliveries at high demand and demonstrates how not meeting the promises effect the business in the short or long run in terms of Churn, Repeat, Net Promoter Score(NPS), etc. We will discuss about the necessity of batching the orders and its advantage and disadvantages on the system. Finally we will go through how we build a Deep Learning system, FOPS, to predict the orders which are going to batch and its effect on the overall environment of the online food delivery space
We will talk about how we solved the problem with the help of Machine Learning/AI, Operations and Engineering.
Content of the talk will be around:
1) Batching Overview.
2) Problems with the existing framework
3) Customer Experience angle in batching
4) First Order batching Prediction
5) How First Order batching Prediction turned the numbers towards Swiggy
6) Next Steps
Senior Data Scientist at Swiggy