A modular, culturally-rooted BPMN 2.0 process automation platform built on Java 21 and Spring Boot 3. Abada provides a lightweight, embeddable workflow engine with first-class observability and production-ready Docker deployment for dev, test, and prod.
For a deep technical overview and deployment architecture, see docs/architecture/overview.md.
Abada is designed for the modern cloud-native stack. Observability is not an add-on; it is woven into the core execution engine.
- OpenTelemetry Native: The engine emits rich OTLP traces, metrics, and logs out of the box. No sidecars or agents required for basic visibility.
- Full-Stack Visibility: Trace every process instance from REST API call → BPMN Element → Database Query, with correlated logs at every step.
- Modern Stack: Integrated with Jaeger (Tracing), Prometheus (Metrics), Loki (Logs), and Grafana (Visualization).
- Actionable Metrics: Pre-configured metrics for process duration, task throughput, and error rates.
- Unified Logging: Centralized log aggregation with automatic trace correlation for debugging distributed workflows.
👉 See the Observability Reference Guide for full configuration details.
Abada is a modular platform composed of:
abada-engine: Core BPMN execution engine (REST API)orun: Monitoring and observability dashboard (frontend)tenda: End-user task UI (frontend)admin: Process administration (frontend)semaflow: Natural-language to BPMN tooling
All components are containerized and deploy independently. For design, boundaries, and runtime topology, see docs/architecture/overview.md.
The easiest way to run the full Abada Platform (Engine, Tenda, Orun, Observability) is using our automated quickstart script. You only need Docker installed.
curl -sSL https://raw.githubusercontent.com/bashizip/abada-engine/main/release/quickstart.sh | bashFor developers contributing to the project:
Hosted with in-memory H2 database, pre-configured users, and full observability tools.
Option 1: Build & Run (Recommended) Builds the engine locally (fast) and launches the complete stack.
./scripts/build-and-run-dev.shOption 2: Start Only Starts the containers if the image is already built.
./scripts/start-dev.shOption 3: Rebuild Engine Only (Fast Iteration) Rebuilds and restarts only the engine service to apply code changes.
./scripts/build-dev.shServices:
- Abada Engine: http://localhost:5601/api
- Abada Tenda (Task UI): http://localhost:5602
- Abada Orun (Monitoring UI): http://localhost:5603
- Grafana: http://localhost:3000 (admin/admin123)
- Jaeger: http://localhost:16686
For production deployment and manual Docker Compose commands, see the detailed Docker Deployment Guide.
For endpoints, authentication, and examples, see docs/development/api.md.
- Core Execution: BPMN parsing and execution; user, service, external, and script tasks; message/signal/timer events; XOR/AND/OR gateways.
- Persistence: H2 (dev/test) and PostgreSQL (prod) with HikariCP pooling.
- Observability: Full OpenTelemetry instrumentation. Tracing for all BPMN elements. Metrics for process/task performance.
- Deployment: Traefik load-balanced, stateless, horizontally scalable engine instances.
- Security: Header-based user context (Role model planned).
See docs/development/roadmap.md for the staged milestones from 0.8.2-alpha to 1.0.0-beta.
Issues and PRs are welcome. Please include tests for any new engine behavior.
MIT