Editorial Outlook: How AI Agents May Reshape Business Process Discovery

The automation field has long been animated by a tension between human judgment and machine execution. Humans identify processes, define scope, and make qualification decisions. Machines execute the automations that humans design. Agentic tooling may shift this division of labor as capabilities and governance mature.

Editorial outlook: The emergence of agentic AI - systems where language models orchestrate complex, multi-step tasks with limited human supervision - is beginning to close the loop. AI agents can now conduct discovery interviews (VARA does this today), generate process maps, produce qualification assessments, and recommend automation investments. A possible next phase includes agents that proactively identify opportunities from process logs, draft and test automation solutions, and monitor deployed automations; timing and capability will vary by platform, process, and governance.

What Agentic AI Means for Process Discovery

Process discovery - the systematic identification and qualification of automation opportunities - has historically required human effort across stages such as designing the intake process, conducting interviews, reviewing findings, and making pattern recommendations. AI has been a tool in this workflow (ML models for pattern classification, NLP for information extraction from interview transcripts), while human review remains important for consequential decisions.

Agentic systems are changing this. A multi-agent discovery architecture might work like this: a discovery orchestration agent continuously monitors process execution data (from ERP, CRM, ticketing systems) and identifies statistical anomalies - unusual variance in cycle times, unexpected exception rates, rework patterns - that suggest automation opportunity. When it identifies a candidate, it dispatches an intake agent (like VARA) to conduct the discovery interview. The interview output is reviewed by a qualification agent that applies the Problem-First framework, generates the assessment, and routes the qualified candidate to the appropriate development queue.

The human's role in this architecture can shift from process execution (conducting interviews, applying frameworks) toward process governance (setting criteria, reviewing edge cases, approving high-value candidates, managing the overall portfolio). This can put human expertise where judgment and oversight are most needed.

Editorial outlook: The Agentic Architecture and Timeline

Current capability: Conversational AI Intake and Qualification

Current capability: AI agents can conduct structured discovery interviews, apply qualification frameworks, produce assessed reports, and recommend automation patterns in supported production contexts. VARA represents this tier. The human remains in the loop for final qualification approval and priority decisions; the degree of intake and assessment automation depends on the implementation.

Possible next stage: Proactive Opportunity Discovery

The next tier - agents that proactively discover opportunities from operational data rather than waiting for human submission - is in active development at several platforms. Celonis's AI-powered opportunity identification, UiPath's Autopilot, and Microsoft's Copilot for Process Mining all represent steps in this direction. Commercial maturity will depend on access to reliable process data, qualification quality, and governance controls; this is an editorial outlook, not a dated market forecast.

Possible later stage: Design and Test Automation

The tier after discovery is design. Agentic systems that can draft automation designs - RPA workflow structures, API integration specifications, BPM workflow maps - from qualified intake reports are in early development. The output will still require expert review, testing, and governance before production use. Timing varies by pattern and risk level, so the sequence matters more than a precise calendar estimate.

Longer-horizon scenario: Supervised Optimization

The longer-term vision - agents that monitor deployed automations, identify performance degradation, diagnose root causes, and propose or implement fixes - is technically feasible and actively being prototyped. Lower-risk, well-monitored environments are likely to mature before high-stakes operational contexts, where approval, rollback, and accountability requirements are materially higher.

What This Means for CoE Teams

Start with the available IntakeOS discovery workflow rather than waiting for every future agent scenario. Interview process teams, inspect qualification and maps, and use report or XML outputs to prepare a reviewable handoff. Treat the article's advanced scenarios as possibilities to evaluate, with availability confirmed separately, not as a promise of autonomous implementation.

The progressive deployment of agentic capabilities doesn't eliminate the CoE - it transforms it. The work that consumes the most CoE capacity today (intake interviews, pattern qualification, business case development) will increasingly be handled by AI agents. The work that will grow in importance (governance design, exception handling strategy, agent oversight, business stakeholder relationships, cross-domain portfolio management) requires the specifically human capabilities that AI augments rather than replaces.

Editorial outlook: CoE leaders planning for agentic automation should ask what decisions should remain with humans, where human judgment adds value, and how to govern agent-driven discovery and deployment responsibly. These questions matter more than any single platform choice because they shape the organizational design of the automation function.

"The question isn't whether AI agents will transform automation discovery. They already are. The question is whether your organization is designing for the world where agents do the discovery and humans do the governance - or still designing for the world where humans do everything."

- Chief Automation Officer, Global Technology Company

The Near-Term Practical Agenda

For practitioners preparing for agentic discovery, build the data and governance infrastructure these systems rely on: clean process event logs, structured intake data, and documented automation patterns and decision criteria. These foundations improve readiness, but they do not guarantee a particular deployment timeline or commercial outcome.

Frequently Asked Questions