Business Process Automation

Part of the worlds-biggest-software-project initiative.

An AI-native, open-source platform for no-code workflow automation, BPMN process governance, and process mining in a single coherent product.

Business Process Automation is a candidate project for an OSS platform that unifies what is today fragmented across Zapier-style integrations, Camunda-style BPMN orchestration, and Celonis-style process mining. It is aimed at operations leaders, IT teams, and process owners who need self-hostable automation with AI agents, audit-grade governance, and process intelligence — without enterprise-tier licensing costs.

Why Business Process Automation?

No open-source platform combines workflow execution, BPMN governance, and process mining. n8n and ActivePieces cover execution; Camunda covers BPMN; Celonis covers mining — none combine all three.
Per-task SaaS pricing penalises complex automations. Zapier at $29.99/mo allows only 750 tasks; complex multi-step automations are economically unviable at volume.
Process mining is locked behind enterprise pricing. Celonis ($100K–$300K+/yr) and UiPath Process Mining are inaccessible to SMBs and mid-market organisations that would benefit most.
n8n's Sustainable Use Licence is not OSI-approved open source and restricts commercial SaaS use, leaving a genuine permissive-licence gap for a workflow platform.
Incumbent AI is shallow. Most tools have bolted natural-language builders onto trigger-action UIs; none combine NL design, agentic execution, process mining, and adaptive optimisation in one architecture.

Key Features

Workflow Authoring & Execution

Visual workflow builder with trigger-action design and conditional branching
Webhook, schedule, HTTP, and email triggers covering core integration patterns
50+ pre-built integrations across CRM, ERP, communication, storage, and ticketing categories
Self-hosting with unlimited executions; no per-task billing ceiling
Role-based access control for workflow creators, editors, and viewers

AI-Native Design & Agents

Natural-language workflow builder: describe an automation in plain language and generate a deployable workflow
AI agentic nodes: autonomous agents that reason about process state, select actions, and execute multi-step tasks
Traceable agent reasoning rendered on the same canvas as deterministic steps for unified debugging
Agentic exception handling: AI agents resolve workflow exceptions within policy guardrails and log decisions for audit

BPMN Process Governance

BPMN 2.0 process modelling for formally governed, long-running business processes
State persistence, retries, timeouts, and SLA tracking on every process instance
DMN 1.3 decision evaluation for business rule logic within process flows
Instance-level monitoring with replay of failed executions for regulated audit evidence

Process Intelligence

Process performance dashboard with execution time, error rate, and bottleneck identification per workflow and step
Process mining to infer actual process behaviour from event log data and surface conformance gaps
AI process discovery from unstructured data (email, documents, chat) without requiring structured ERP event logs
Adaptive process optimisation: ML-based recommendations for workflow modifications based on observed performance

Embedding & Extensibility

Embedded automation SDK for SaaS developers to offer white-label automation inside their own products
HTTP node for connecting any REST API beyond pre-built integrations
Code steps for custom JavaScript/Python logic where pre-built nodes are insufficient

AI-Native Advantage

Incumbent BPA tools execute fixed processes designed by humans against structured data. This project treats AI as a first-class layer: natural-language process design lowers the skill barrier below trigger-action logic; AI agents handle exceptions autonomously inside formal policy guardrails with audit trails; process mining infers process models from email, chat, and documents — opening discovery to organisations without clean ERP event logs; and adaptive optimisation continuously suggests workflow modifications from observed performance data.

Tech Stack & Deployment

The project targets self-hosted, cloud, and embedded deployment modes. It aligns with established standards — BPMN 2.0 for process modelling, DMN 1.3 for decision logic, CMMN for case management, IEEE 1849 (XES) for process-mining event logs, OpenAPI/REST for integrations, and CloudEvents for event-driven interoperability. A clean architectural path is to build on Camunda's Apache 2.0 Zeebe execution engine for BPMN, layered with a new permissively-licensed AI design and monitoring surface and an integration layer modelled on the ActivePieces (MIT) approach.

Market Context

The global workflow automation market is valued at approximately $26.01B in 2026 (Mordor Intelligence), with CAGR estimates of 9.41% through 2031 to 11.2% through 2034 (Fortune Business Insights). Pricing spans enterprise BPM suites at $50K–$500K+/yr (Appian, Camunda Enterprise), mid-market tools at $1,500–$20K/mo (Power Automate, Kissflow), SMB SaaS at $0–$500/mo (Zapier, Make), and process mining at $100K–$300K+/yr (Celonis, UiPath Process Mining). Primary buyers are Heads of Operations / COOs, IT and digital transformation leads, business analysts and process owners, and CFOs evaluating automation ROI.

Project Status

This project is in the research and specification phase.
Contributions, feedback, and domain expertise are welcome.

Contributing

We welcome contributions from developers, domain experts, and potential users. See CONTRIBUTING.md for guidelines.

Important: All contributions must be your own original work or clearly attributed open-source material with a compatible licence. Copyright infringement and licence violations will not be tolerated and will result in immediate removal of the offending contribution. If you are unsure whether a piece of code, text, or other material is safe to contribute, open an issue and ask before submitting.

Licence

Licence to be determined. Apache 2.0 or MIT are the leading candidates, both of which would avoid the commercial-use restrictions of n8n's Sustainable Use Licence and remain compatible with building on Camunda's Apache 2.0 Zeebe engine.