Flowable

In October 2016, the lead developers of Activiti (software) left Alfresco (software) and started the Flowable Open Source project based on a fork of Activiti code.^[3][4]

The first version of Flowable was 5.22, based on a fork of Activiti 5.21, but added Transient Variables.^[5] The first release of Flowable version 6.0^[6] was based on a fork of Activiti version 6 beta 4. Version 6 of the Flowable engine includes a rewrite of the core process virtual machine.

Flowable has a modular, model-driven platform for business process and case automation that supports both low-code and traditional code development. Its architecture centers on runtime engines that execute workflow, case, and decision models defined using open standards.These engines can be embedded within Java applications or deployed as standalone services, and are typically accessed through Java or REST‑based APIs.

In addition to its core execution engines, the platform includes supporting components for model design, identity management, deployment, and operational monitoring. It supports both code‑centric and low‑code development approaches, allowing automation logic to be expressed programmatically or through graphical models, depending on the use case and deployment context.

The platform comprises:

• Flowable Work^[7]: A low-code platform for managing structured workflows and tasks, providing features for process design, task management, and integration with external systems.
• Flowable Design^[8][9]: The central modeling environment for creating process, case, decision, and form models. It supports drag-and-drop interfaces, reusable templates, and Git-based versioning.
• Flowable AI Studio^[10]: A separately licensed module for AI-assisted automation and agent-based orchestration. Key features include:  
  • Multi-agent orchestration^[11][12][13]: It includes a dedicated agent engine that operates alongside BPMN and CMMN engines to manage specialized agents for tasks such as documentation, verification, and decision-making.
  • LLM and RAG integration: Supports integration with commercial and private large language models (LLMs) and Retrieval-Augmented Generation (RAG) for information retrieval within workflows.  
  • Generative modeling^[14]: Provides functionality for creating workflows and forms from natural language input.  
  • Content analysis: Agents are used for tasks such as document summarization, sentiment analysis, and data extraction from unstructured documents.
  • Audit and governance^[15]: Includes audit logs and governance mechanisms that record agent actions and automated decisions.
• Flowable Control: A tool for monitoring and managing Flowable’s process and case engines. It provides tools to identify issues, correct case data, and maintain process execution. You also get detailed runtime engine insights for troubleshooting and performance monitoring.
• Flowable Hub^[16]: The central management application in Flowable Cloud, organized into Management and Environment areas. It lets you configure users and permissions, manage runtime environments, deploy applications, and review runtime or historical data, similar to Flowable Control for on-premise deployments.

Flowable added AI‑assisted automation capabilities, reflecting a broader shift in business process platforms toward agent‑based and adaptive automation known as an Agentic Case Platform.^[17] This means integrating AI agents directly into BPMN and CMMN workflows. Collaboration between human users, traditional systems, and AI agents is a result of this approach. Supported capabilities include:

• Specialized agent types such as:  
  • Document agents^[18] for data extraction and classification.
  • Knowledge agents^[19] for RAG-based querying.
  • Orchestration agents^[20] for coordinating other agents and guiding process flow.
• Governance and auditability of agent actions^[21].
• Orchestration of processes and cases using BPMN and CMMN models.

Flowable’s core capabilities span workflow architecture, open standards compliance, and advanced AI orchestration. The platform includes the following features:^{[22] [23]}

• AI-assisted modeling: Supports AI-assisted modeling to create workflows, forms, and cases from natural language descriptions.
• AI-powered case management: It offers real-time agent assistance for summarization, decision support, and autonomous task execution.
• Process orchestration: End-to-end coordination, automation, and management of all elements involved in completing a work task, from people and workflows to technology.
• Content analysis and generation: Provides tools for sentiment analysis, document summarization, and content translation.
• Integration and external agent connectivity: Supports integration with external systems through REST APIs, web services, event listeners, and connectors for services such as AWS Bedrock, Azure OpenAI, Salesforce Agentforce, and RPA tools.
• Security and scalability: Designed for enterprise deployments with high availability and built-in security and compliance mechanisms.
• Open Standards: Implements the BPMN, CMMN, and DMN standards for workflow, case, and decision modeling.
• Composable architecture: Uses a modular architecture that allows components and models to be reused across applications and workflows.  
• Enterprise data control: Features such as Data Objects enable management of data models and mapping between internal and external data sources.
• Auditability and compliance: Maintains process history and audit logs that provide traceability of automated actions and decisions.

• Flowable 2025.1 Release^[24][25]: Introduced a dedicated agent engine alongside the BPMN and CMMN engines, supporting multi-agent collaboration, external agent orchestration, and AI-assisted case views.
• Flowable 2025.2 Release^[26][27]: Expanded governed agentic automation with enterprise‑grade multi‑agent orchestration, A2A‑compatible external agent support, and enhanced AI Studio capabilities, adding deep runtime visibility, AI-assisted design and modeling, and stronger governance for regulated environments.
• Industry recognition: Featured in the 2025 Gartner® Business Orchestration and Automation Technologies report (BOAT)^{[28] [29]}and Forrester Wave™ for Digital Process Automation (DPA)^[30][31], recognized for orchestration precision and governance-first AI capabilities.
• Enterprise adoption: Reported adoption includes organizations such as UBS and Deutsche Post^[32] for document processing and workflow management.