From Dream to Digital: Chowman’s Food Delivery Journey

High Server Load

Chowman processes a high number of orders daily, which causes performance issues during traffic hours. This leads to order synchronisation problems with Pet Pooja, and server load during peak order processing must be controlled.

We set up a load balancer to distribute incoming traffic among various servers. This reduced server burden during peak hours and ensured that if one server went down, others could take over — increasing system availability and reliability.

Lost Chat Images

While using the support chat, some images were lost during customer interactions. This impacted communication quality and played an important role in user satisfaction.

Chat images were previously saved on local servers, causing issues when load balancing spun up new sub-servers under high load. We moved image storage to AWS S3 buckets so all images are safely stored and accessible regardless of which server is active.

Database Load Management

Frequent read-and-write operations on the database caused performance issues in data handling. Database performance was critical to the overall speed of the application.

We selected Amazon Aurora, which enables read replicas to balance database load. This increases read/write efficiency and allows a new primary to be elected from a read replica without significant service disruption. API optimisation reduced the frequency of database queries.

Fraud Prevention & Security

Several security issues were encountered: discount coupons were automatically applied without proper verification, and users could place orders without completing online payments. These flaws highlighted the need for stronger transaction authenticity to prevent fraud.

We added server-side validation to improve security — validating all critical data (prices, quantities, transaction parameters) on the server rather than the client side. Only transactions authorised by the server are processed, significantly reducing fraud risk.