Uber Backend Clone

• Architected a production-grade ride-hailing backend clone applying Clean Architecture, Domain-Driven

Design (DDD), and CQRS to enforce strict separation of concerns across Domain, Application, Infrastructure,

and API layers

• Designed bounded contexts and rich domain aggregates (Trip, Driver, Rider, Payment) with domain events

driving cross-context communication via Kafka

• Implemented CQRS with separate read/write models — EF Core for transactional writes, Dapper for optimised

read queries — and MediatR for decoupled command/query dispatch

• Built real-time ride matching and driver location tracking using geospatial queries on Redis Stack, with surge

pricing and fare calculation logic encapsulated in the domain layer

• Leveraged Kafka for event streaming across the trip lifecycle (ride requests, driver updates, payments,

notifications) with background IHostedService consumers per domain

• Integrated Keycloak for authentication and role-based authorisation (Rider/Driver/Admin) with JWT

propagation across services

• Deployed with Docker and Kubernetes; observability via ELK stack (Elasticsearch, Logstash, Kibana), Elastic

APM for distributed tracing, and Metricbeat for infrastructure metrics

• Tools Used: .NET, PostgreSQL, Redis Stack, Kafka,