Architected and developed a scalable backend platform simulating a social commerce ecosystem (#EcommercePlatforms, #ScalableSystems), integrating user profiles, product catalogs, purchase workflows, and a basic ML-based recommendation engine. Designed with Python (FastAPI), PostgreSQL, and AWS services to achieve sub-second query performance and horizontal scalability across regions. Incorporated exponential backoff and API rate-limiting strategies to maintain stability during high-load conditions. Led system design with an emphasis on modular growth and real-time user engagement features.
Designed a real-time logistics event processing system (#SupplyChainTechnology, #EventDrivenArchitecture) to track shipments, warehouse movements, and customer notifications. Implemented with Kafka, Python, and DynamoDB to support high-throughput, low-latency operations across multiple geographies. Delivered a merchant-facing dashboard for real-time visibility, enhancing platform reliability and operational transparency.
Live-Streaming Commerce Platform MVP
Built a lightweight live-commerce platform prototype (#ContentCommerce, #LowLatencySystems), integrating live video, interactive chat, and instant purchase capabilities during sessions. Developed a FastAPI-based backend with WebSocket integration and a mobile-responsive frontend, ensuring real-time interactions and resilient session management during peak concurrent events.
Designed and launched a fully serverless product recommendation API (#ServerlessArchitecture, #PersonalizationSystems), triggered dynamically by user behavior events. Built using Python-based AWS Lambda functions, Step Functions, and EventBridge to orchestrate scalable, cost-optimized compute pipelines. Integrated observability, latency tracking, and retry logic using jittered exponential backoff to ensure reliability under burst traffic. Applied regional failover strategies and monitored performance with AWS X-Ray and CloudWatch metrics.
Global Supply Chain Visualization Dashboard
Engineered a real-time visualization dashboard (#GlobalLogistics, #RealTimeMonitoring) tracking shipment movements, inventory status, and delivery flows across regions. Integrated real-time event ingestion from Kafka into a dynamic TypeScript-based frontend, with Python-based ETL components aggregating logistics data. Applied latency analysis techniques to tune map refresh cycles and ensure sub-second update frequencies. Designed to support both batch-mode history views and real-time mode for live operations monitoring.
