Case study
Swiggy
Food & quick-commerce delivery
Overview
Swiggy scaled from restaurant delivery into quick commerce (dark stores, SLAs in minutes). That shifts the problem from batch deliveries to tight inventory + picker + rider coordination.
Students can think of the stack as real-time ops: ETA promises are marketing; fulfilling them is systems engineering plus on-ground playbooks.
Technical problems at scale
Inventory sync and overselling
When SKUs sell in-store and online, stock counts race. Optimistic locking, reservation holds, and async reconciliation prevent selling items that cannot be picked.
Last-mile density and batching
In cities, batching multiple drops per rider improves unit economics but complicates ETAs. Heat maps and surge pricing feed back into rider incentives.
Partner apps and offline modes
Restaurant tablets and rider phones may be on flaky networks. Offline-first UX, sync queues, and conflict resolution mirror mobile distributed systems case studies.
Experimentation under network effects
A/B tests in marketplaces spill across users and zones. Experiment design must avoid double-dipping riders and skewing control cities.
Systems & patterns you will hear about
- Mobile + edge-friendly APIs
- Inventory OLTP
- Real-time location streams
- Feature flags & experimentation
- Pricing / surge engines
Case-study angles
Map a happy path vs degraded path for “Instamart order in 10 minutes.” What do you show the user if the picker queue is full but payment succeeded?
Role-play an incident: spike in cancellations—which dashboards and traces would you open first?