Maestro user guide

Maestro supports two orchestration models: Maestro BPMN (sequential, BPMN-based workflows) and Maestro Case (stage-based, goal-driven orchestration for case management). Choose Maestro BPMN when work follows a predictable, repeatable sequence every time. Choose Maestro Case when work is long-running, exception-heavy, and requires dynamic routing between stages based on evolving data and human judgment.

This document provides a decision framework to help automation architects, business analysts, and solution engineers determine which model fits a given business scenario — or whether both should work together.

Audience: Intermediate to Advanced — Automation Architects, Solution Engineers, Business Analysts

Traditional process orchestration excels when the flow of work is predictable and repeatable. A purchase order approval, a scheduled data extraction, or a notification pipeline follows the same path every time. Maestro BPMN models these flows efficiently.

However, many business scenarios do not follow a single, predetermined path. They may span days or weeks, involve multiple teams, and require decisions that depend on what just happened — not on a fixed sequence defined at design time. In these situations, exceptions are not rare — they are expected.

Consider an insurance claim: it starts with intake, moves to investigation, then assessment, then settlement. But along the way, an adjuster might request additional documentation, sending the case back to an earlier phase. A fraud signal might terminate the process mid-stream. A customer might withdraw the claim at any point. No single flowchart can capture every path.

These scenarios need an orchestration model that provides structure without rigidity — one that allows automation and AI to handle routine work while enabling humans to step in when judgment or policy decisions are required.

Maestro offers two distinct models for orchestrating work. Each model addresses a different class of problem.

Maestro BPMN is a BPMN-based workflow that defines a specific sequence of steps. It runs from start to finish along a path determined at design time. Maestro BPMN supports branching, parallel execution, and human-in-the-loop tasks, but the overall structure is a directed flow.

Use Maestro BPMN when:

Maestro Case orchestrates long-lived, goal-driven work about a specific situation — a case. A case holds data, policies, tasks, and history to drive an auditable outcome (for example, a refund, a claim decision, or an investigation closure). Instead of a fixed sequence, a case plan defines stages (named phases) and the rules that govern transitions between them. The actual path through the plan is determined dynamically at runtime based on case data.

Use Maestro Case when:

The following comparison highlights the structural and behavioral differences between the two models. Use it to evaluate which model matches the characteristics of your scenario.

Case management adds the most value when work is long-running, exception-heavy, and human-centric. It allows automation and agents to handle routine tasks while ensuring that humans are involved at the right moments. This approach reduces delays caused by manual follow-ups and helps teams maintain visibility and control across the entire lifecycle of a case.

The following characteristics signal that case management is the right choice:

Not every process needs case management. If a process is short-lived, predictable, and follows the same sequence every time, Maestro BPMN is simpler and more efficient. Introducing case management in those situations adds unnecessary complexity.

Stick with Maestro BPMN when:

Examples: invoice data extraction, scheduled report generation, single-approval purchase orders, notification pipelines, and batch data migrations.

Traditional case management relies on human knowledge workers to make routing decisions, handle exceptions, and drive cases forward. Maestro Case replaces that bottleneck with AI agents at two levels:

The Case Manager is the event-driven orchestrator of a case. It uses two complementary methods: rules first (deterministic CMMN rules from the case plan resolve the high-volume happy paths) and agent reasoning as fallback (an LLM-based Case Manager Agent reasons over case state, the case plan, and configured policies to handle gaps, exceptions, and judgment calls). If neither rules nor the agent can resolve a decision — due to ambiguous data, conflicting policies, or a scenario outside the agent's authority — the Case Manager escalates to a human.

This hybrid approach provides both reliability (deterministic rules for every scenario you can anticipate) and flexibility (AI reasoning for everything rules do not cover).

Individual tasks within a stage can be assigned to AI agents — alongside human tasks, API calls, RPA bots, and connectors. An agent task might categorize expenses, flag anomalies, draft a response, or extract data from unstructured documents. Maestro supports both UiPath Agents and External Agents for these tasks.

Agents are the mechanism that powers faster resolution, higher throughput, and fewer manual touches. The value proposition is the business outcome: streamlined exception-heavy processes, improved KPIs, and end-to-end visibility. Agents are an input to the solution, not the output.

Maestro BPMN and Maestro Case are not mutually exclusive. A case task can invoke a Maestro BPMN via the Maestro Agentic Process task type — the BPMN runs with its own orchestration logic and returns a result to the parent case. This pattern is useful when a specific stage within a case contains a well-defined, linear subprocess (for example, a multi-step audit workflow) that benefits from BPMN modeling. The reverse is also supported: a Maestro BPMN can invoke a Maestro Case as one of its task types.

Similarly, a case can spawn a Child Case — another case definition with its own lifecycle, stages, and tasks — linked to the parent via case ID. This supports scenarios where a parent case (such as a claim) needs to trigger a subordinate investigation with its own independent progression.

For teams evaluating case management, the following core concepts define the model:

For detailed definitions of all constructs, refer to the Maestro Case core concepts reference.