Context & Engineering Constraints
Dubizzle serves millions of users across the MENA region. The platform needed ultra-low-latency event streaming for lead tracking, automated dealer onboarding, and modernization of core legacy services.
Architecture & System Approach
- 01Engineered distributed asynchronous data pipelines and microservices using Python, Django, Celery, and Apache Kafka.
- 02Automated the enterprise B2B dealer onboarding and transaction pipeline with Checkout payment gateway integration.
- 03Decomposed legacy monoliths into containerized microservices running on AWS Kubernetes (EKS).
- 04Optimized Redis caching layers and PostgreSQL query indexing for sub-100ms response times at peak load.
Measurable Business & Technical Outcomes
- Cut enterprise client onboarding latency by 90%, significantly accelerating commercial partner velocity.
- Handled high-throughput real-time lead streaming with zero data loss across millions of marketplace events.
- Reduced database read load by 65% through Redis query caching.
“Production reliability is not an afterthought—it is a property of the entire system boundary: typed schemas, explicit state machines, durable recovery, and deterministic human oversight.”