Opinionated starter for building data-processing microservices with job orchestration, data enrichment, quality gates, lineage tracking, and event-driven capabilities.

• Overview
• Features
• Requirements
• Installation
• Quick Start
• Configuration
• Documentation
• Contributing
• License

Firefly Framework Starter Data provides a production-ready architecture for building data-processing microservices. It combines job orchestration for batch and async data processing with a data enrichment framework for third-party provider integration, all built on hexagonal architecture principles.

This starter is designed for data-tier microservices that need to:

• Execute complex, orchestrated data processing workflows
• Integrate with external data providers (credit bureaus, financial data, business intelligence)
• Enforce data quality rules and track data lineage across enrichment pipelines
• Implement CQRS patterns for command/query separation
• Build event-driven architectures with automatic event publishing
• Support distributed transactions using Saga, TCC, and Workflow patterns

Job Orchestration

• Async and sync execution modes with abstract controller and service base classes
• Standardized RESTful endpoints with job execution tracking and audit trails
• Configurable timeout enforcement per job stage

Data Enrichment

• Pluggable DataEnricher and EnricherOperation framework with tenant isolation
• Enrichment discovery controller for runtime operation catalog
• Smart enrichment controller with strategy-based routing
• Fallback chains with primary/secondary provider failover
• Per-provider resiliency configuration (circuit breaker, retry, rate limiter, bulkhead)
• Enrichment caching with configurable key generation and TTL
• Cost tracking and estimation per enrichment request

Data Quality

• Rule-based validation engine with configurable severity levels
• Fail-fast and collect-all evaluation strategies
• Built-in rules for null checks, range validation, pattern matching, and custom logic
• Quality gate integration for enrichment pipelines

Data Lineage

• Automatic lineage tracking across enrichment operations
• Pluggable LineageTracker with in-memory default implementation
• Lineage records with source, destination, timestamp, and metadata

Data Transformation

• Reactive DataTransformer interface with TransformationChain composition
• Built-in transformers for field mapping and computed fields
• Custom transformer support via functional interface

Infrastructure

• CQRS auto-configuration for command/query integration
• EDA auto-configuration for event-driven data processing
• Orchestration engine integration (Saga, TCC, Workflow)
• Observability with Micrometer metrics and distributed tracing
• Persistence auto-configuration for job results and audit entries

• Java 25
• Spring Boot 3.x
• Maven 3.9+
• PostgreSQL database (for job persistence)

<dependency>
    <groupId>org.fireflyframework</groupId>
    <artifactId>fireflyframework-starter-data</artifactId>
    <version>26.02.07</version>
</dependency>

import org.fireflyframework.data.service.AbstractResilientDataJobService;
import org.fireflyframework.data.controller.AbstractDataJobController;

@RestController
@RequestMapping("/api/data-jobs")
public class MyDataJobController extends AbstractDataJobController<MyRequest, MyResponse> {

    public MyDataJobController(DataJobService<MyRequest, MyResponse> service) {
        super(service);
    }
}

@Service
public class MyDataJobService extends AbstractResilientDataJobService {

    @Override
    protected Mono<JobStageResponse> doStartJob(JobStageRequest request) {
        return processData(request)
            .map(result -> JobStageResponse.success(
                JobStage.START, request.getExecutionId(), "Job completed"));
    }

    @Override
    protected Mono<JobStageResponse> doCheckJob(JobStageRequest request) {
        return Mono.just(JobStageResponse.success(
            JobStage.CHECK, request.getExecutionId(), "Job is running"));
    }

    @Override
    protected Mono<JobStageResponse> doCollectJobResults(JobStageRequest request) {
        return Mono.just(JobStageResponse.success(
            JobStage.COLLECT, request.getExecutionId(), "Results collected"));
    }

    @Override
    protected Mono<JobStageResponse> doGetJobResult(JobStageRequest request) {
        return Mono.just(JobStageResponse.success(
            JobStage.RESULT, request.getExecutionId(), "Result retrieved"));
    }

    @Override
    protected Mono<JobStageResponse> doStopJob(JobStageRequest request, String reason) {
        return Mono.just(JobStageResponse.success(
            JobStage.STOP, request.getExecutionId(), "Job stopped"));
    }
}

firefly:
  data:
    job-orchestration:
      enabled: true
      max-concurrent-jobs: 10
    enrichment:
      cache-enabled: true
      cache-ttl-seconds: 300
      operations:
        cache-enabled: true
        cache-ttl-seconds: 600
    quality:
      enabled: true
      fail-fast: false
    lineage:
      enabled: true
    persistence:
      enabled: true
    observability:
      metrics-enabled: true
      tracing-enabled: true

Comprehensive documentation is available in the docs/ directory:

• Documentation Index - Full table of contents and guide overview
• Architecture - Hexagonal architecture and design patterns
• Getting Started - Step-by-step setup guide
• Configuration - All configuration properties
• Data Enrichers Guide - Building enrichers and operations
• Data Jobs Guide - Job orchestration and lifecycle
• Data Quality - Quality rules and validation engine
• Data Lineage - Lineage tracking across pipelines
• Data Transformation - Transformation chains and field mapping
• Resiliency - Circuit breaker, retry, and per-provider configuration
• API Reference - REST endpoint specifications
• GenAI Bridge - Integration with fireflyframework-genai

Contributions are welcome. Please read the CONTRIBUTING.md guide for details on our code of conduct, development process, and how to submit pull requests.

Copyright 2024-2026 Firefly Software Solutions Inc.

Licensed under the Apache License, Version 2.0. See LICENSE for details.