Kafka, in motion
A live simulation of the event-streaming patterns behind my Global Freight platform, told through one shipment's journey from order to doorstep.
Global Freight & Shipment Platform
2024 - 2025The Kafka build this demo is modelled on
- shipment events through Kafka
- 50k+
- lower API response latency
- 40%
Spring Boot microservices for shipment tracking, bidding and logistics, talking to each other through Kafka.
Avro schema contracts across the service topology, with CQRS and Event Sourcing for auditable workflow state.
SAGA orchestration with compensating transactions keeps data consistent when a step fails part-way.
Java / Spring Boot / Kafka / Avro / Axon / MongoDB / Redis / Docker
DelayLens
2026The supply-chain data behind the demo
- late-delivery rate, root-caused
- 54.8%
- late after revised delivery promises
- 18.9%
Hadoop ETL over 180,519 DataCo order lines, transformed by 33 MapReduce jobs into order-level delivery KPIs.
First Class shipments ran 95.3% late, which is why the simulation's First Class orders run late most often.
Python / Hadoop / MongoDB / LangGraph / Next.js
Asset Marketplace
2021 - PresentEvent sourcing and CQRS
Domain-driven design with CQRS and Event Sourcing on EventStoreDB, inside an NX monorepo.
GraphQL APIs and subscriptions with payment workflows (Stripe, Coinbase, BitPay).
TypeScript / NestJS / EventStoreDB / GraphQL / MongoDB
The simulation is illustrative. First Class orders run late most often, as DelayLens found in the DataCo supply-chain data.
