Agentic Business Process Management: A Research Manifesto

This is a research manifesto presenting conceptual foundations rather than implemented software or validated algorithms. At 19 days old with 0 stars and no forks (the 18 forks appear to be repo clones rather than active development), this is a pre-publication positioning paper. The manifesto articulates a novel combination of agent-oriented computing with traditional BPM, addressing a real gap in how to govern autonomous systems in enterprise contexts. However, defensibility is minimal because: (1) it exists only as a theoretical framework with no reference implementation, (2) major platforms (AWS, Azure, Google Cloud) and workflow orchestration vendors (Salesforce, ServiceNow, SAP) are actively investing in agentic AI workflows and process automation, (3) the domain is nascent but high-value, making acquisition or internal development by incumbents highly likely once the market forms, (4) the paper itself is novel in positioning but not technically defensible—competitors can read the manifesto and implement similar concepts. The 'high' platform domination risk reflects that Microsoft (Copilot for BPM, Dynamics integration), SAP (Signavio + agentic extensions), and OpenAI/Anthropic (agent orchestration frameworks) all have direct incentive to absorb this capability within their workflow platforms. Medium market consolidation risk reflects that specialized BPM vendors (Camunda, Celonis) could also implement agentic governance, but they're smaller and slower to innovate. The displacement horizon of 1-2 years assumes platform vendors will prototype agentic BPM capabilities within this window once initial adoption signals emerge. This is a thought-leadership contribution that may influence the field but will not create a defensible product moat without substantial engineering and ecosystem building beyond the manifesto itself.