v3.0.0 - Revolut integration and refactoring

This release introduces full Revolut integration into the TrStream pipeline and includes a strategic refactoring to reduce code duplication and improve reuse across services.

Highlights

• Revolut Integration Stack

  • Hybrid architecture combining local components (generator, worker) with publicly exposed services (webhook, event queue)
  • Integrated Sandbox Revolut API for authentication and transaction event testing
  • Introduced a custom CLI to facilitate authentication, webhook registration and account listing

• Services Refactoring

  • Defined base modules to centralize shared components across services (reader/writer, producer/consumer, configuration, logging)
  • Refactored configuration structure and updated YAML definitions
  • Removed relative imports to improve module clarity and packaging consistency
  • Standardized environment variable ingestion within Docker images

v2.0.0 - Stripe integration and storage update

This release introduces the full Stripe integration in the TrStream pipeline, along with key storage and processing updates.

Highlights

• Stripe Integration stack
  • Fully containerized webhook and event generator services
  • Host-based Stripe CLI event forwarding for real and test events
  • Automated scripts for starting and stopping the stack
• Storage and processing updates
  • Updated MinIO paths and partitioning logic
  • Event normalization for Stripe and internal producers
  • Introduction of medallion architecture: raw → processed → analytics
• Service adaptations
  • Partitioner evolved into a processor and compacter slightly adapted to handle the new storage layout
  • Schema- and source-aware processing maintained for query-ready outputs

This release enables local testing of Stripe events, improves data organization and lays the groundwork for further integrations (e.g. Revolut API, Airflow scheduling)

v1.0.0 - Stable pipeline architecture

This release marks the first stable version of TrStream, a distributed real-time transaction processing pipeline.

Highlights

• End-to-end streaming pipeline built on Kafka and S3-compatible storage
• Horizontally scalable producers and consumers
• Immutable Parquet-based data lake with explicit lifecycle stages
• Dedicated partitioning and compaction services for analytical workloads
• SQL query layer powered by DuckDB with a Streamlit-based editor
• Fully containerized local environment with Docker Compose
• Kafka UI integration for observability

Architecture

The pipeline models a modern fintech-style data flow:

• event-driven ingestion
• durable raw event storage for auditing
• lakehouse-style optimization for analytics

Tooling

• Helper scripts for build, run, scaling and shutdown
• Health checks and deterministic startup order
• Clear separation of responsibilities across services

Status

This release is intended as a stable foundation.
Future work will focus on:

• integration with real transaction APIs (e.g. Stripe, Revolut)
• metrics and observability improvements
• advanced analytics and fraud detection use cases