Logistics and Data Science

Question 1

Learnings

Answer 1

Forecasting: Ensure the right amount of couriers at any moment in the most automatic way
Long/medium-term forecast
Optimisation
Linear programming
Network theory

Matching: To make the best dispatching decisions and provide innovative solutions that support all Glovo users
Real-time fleet allocation
Optimisation
Simulation
Assigned problems

Routing: Provide accurate, reliable, and business-driven estimations about the logistics of every order
Short term forecast
Real-time models
GPS optimization
Reinforcement learning

Question 2

Balancing Supply and Demand

Answer 2

Slot Model: Couriers book in advance when they want to work
Free model: couriers are free to work whenever they want
The number of couriers needed depends on the orders volume → Forecasting
The assignment problem order/courier is not 1:1 → Optimization
Overall we’re seeking an optimal balance
Undercasting has a direct impact on the customer experience
Overcasting is bad for couriers