Orchestrating the agentic enterprise: what’s new in UiPath 2025.10

Simply deploying new technology doesn’t guarantee efficiency. Productivity truly emerges when organizations restructure how work is organized, governed, and measured around that technology. Every technological leap follows the same pattern. Factories didn’t achieve productivity from electricity until they reorganized around distributed power. Early computer adopters didn’t gain efficiency until they redesigned processes to harness digital information flow. AI now stands at the same inflection point. Its promise lies not in automating isolated tasks but in restructuring entire workflows around intelligence and data flow.

Take customer onboarding in financial services. A single process might involve document extraction, background checks, credit scoring, risk assessment, and human review. Each step may be automated or AI-enabled, yet if they operate independently—without shared context, visibility, or governance—the process still breaks. The inefficiency simply moves faster.

That’s where agentic orchestration proves essential. It turns complex, cross-silo processes into coordinated systems that can adapt, learn, and perform consistently. It’s how customer onboarding is completed in minutes instead of hours, how supply chains reroute themselves in real time, and how compliance checks run continuously rather than retrospectively. Without orchestration, enterprises risk repeating the productivity paradox: massive investment, minimal transformation.

Maestro, the orchestration engine for agentic automation

We introduced UiPath Maestro™ earlier this year to give enterprises a way to turn disjointed automation and AI efforts into cohesive, outcome-driven workflows. Since then, customer feedback and adoption have shaped its next stage of prioritization: focusing on how process orchestration can deliver visibility and control not only to IT, but also to business users and process owners.

Case Management and Process Apps: Bringing orchestration closer to operations

At FUSION 2025, we unveiled new Case Management and Process Apps in Maestro for business users and process owners who need to better understand their operations, resolve issues faster, and make data-driven decisions with confidence.

Case Management brings the power of agentic orchestration to complex, case-driven processes such as claims, onboarding, or lending. It gives business and IT teams a single framework to model, manage, and monitor cases end to end.

Each case becomes a living workflow with its own timeline, participants, and defined rules for action. Teams can define stages, assign responsibilities, set service-level agreements (SLAs), and establish escalation rules. Maestro then coordinates the work, triggering collaborative groups of robots and AI agents to execute tasks, apply judgement, surface insights, and track progress at every stage.

For example, in insurance claims, Maestro orchestrates robots and AI agents to automatically validate claim data, check policy coverage, initiate payments for straightforward cases, and escalate ambiguous ones to human review, all while maintaining full visibility and auditability. Managers can use this visibility to monitor active cases, identify bottlenecks, and optimize performance in real time.

Process Apps bring orchestration to life for business teams through intuitive dashboards that display live process data, approvals, SLA tracking, and exceptions requiring attention—all in one view.

A credit officer can approve loans directly from an app powered by Maestro; an HR leader can track onboarding progress across departments; a customer service manager can see which escalations are at risk of breaching SLAs and trigger mitigation actions instantly.

Out-of-the-box Process Apps cover common enterprise workflows such as claims handling, onboarding, and purchase approvals, while the Process Apps SDK enable developers to build custom apps that reflect an organization’s unique processes, governance policies, and brand. Together, Case Management and Process Apps in Maestro extend orchestration beyond IT and automation teams into daily business operations. They give organizations a shared view of how work happens, where it can improve, and how decisions impact performance. In doing so, they make orchestration not just a backend capability, but a frontline advantage.

These capabilities are progressing through preview with early customers and will gain new features and enhancements as they advance towards general availability. With the UiPath 2025.10 release, Maestro also becomes smarter, faster to design in, and better connected to the systems and AI platforms enterprises already rely on.

Smarter design with Autopilot and Maestro

Modeling enterprise workflows has traditionally been slow and technical. Teams would spend weeks mapping processes, translating business logic into BPMN diagrams, and aligning IT with operations before anything could run in production.

UiPath Autopilot™ for Maestro changes that. It brings natural language and generative design directly into process modeling. Teams can describe a workflow in plain English, or upload existing diagrams or documents, and Autopilot instantly creates an executable process model.

This makes process design iterative and collaborative. Business users can focus on intent: “how should this work?”, while Autopilot handles syntax and structure. Developers and analysts can then refine logic, rules, and data connections in real time, making process orchestration faster, easier and more inclusive.

A procurement team can use Autopilot to rebuild its purchase order workflow. The process often spans multiple systems and requires weeks of documentation and configuration. With Autopilot, the model can be generated in days instead of weeks, ready for testing, complete with exception paths and escalation logic.

Faster process modeling with curated, industry-aligned templates

Every organization runs a handful of mission-critical processes that follow similar patterns: invoice handling, supplier onboarding, customer approvals, loan processing. Starting from scratch each time wastes effort and slows transformation.

The new Maestro templates solve that problem. They provide prebuilt, configurable blueprints for common workflows, embedding best practices for structure, logic, and governance. Teams can start with a proven foundation, adapt it to their data and business rules, and go live in days instead of weeks.

This helps enterprises scale orchestration consistently, ensuring processes run with the same compliance, visibility, and performance standards across teams and functions. There are 30 curated BPMN templates across industries, from banking and insurance to healthcare and telecom. Each is aligned with frameworks such as APQC, BIAN, and eTOM. You'll find the templates directly in Studio Web.

Orchestration for any AI agent

Most enterprises aren’t starting from scratch with AI. Across every organization, different teams are already experimenting. Finance may be testing forecasting models, service teams are deploying conversational agents, and data scientists are fine-tuning models in Databricks, while using data stored in Snowflake.

With Maestro, customers can bring these experiments together under unified governance and observability. Maestro now orchestrates AI agents from major platforms like Google Vertex, Microsoft Copilot, Databricks, NVIDIA, Snowflake, Salesforce, and more.

Through Model Context Protocol (MCP) and built-in connectors, enterprises can operationalize any AI model or agent directly in Maestro while maintaining visibility, auditability, and control. This gives organizations tangible advantages:

• Use the right agent for the right challenge. Finance might run a Copilot agent for reporting while operations triggers a Vertex agent for demand prediction. Maestro coordinates them within one process, so each contributes its expertise where it adds the most value.

• Apply consistent security and governance. Connections to external agents are explicitly defined and limited to the actions you authorize. Every data exchange is logged and governed by the same role-based access controls (RBAC) that protect UiPath automations. Agents are deployed into Orchestrator folders that inherit these permissions and RBAC policies, ensuring consistent access control and governance.

• Ensure clear, reliable communication with every agent. MCP not only standardizes how models and agents connect—it provides a shared protocol that defines what can be invoked and how. This structured interaction layer makes agent communication predictable, auditable, and less error-prone at runtime, improving reliability and consistency across different AI systems.

• Correlate every decision across platforms. Maestro captures a unified view of what each agent is doing, in which process, and at what stage, enabling cross-platform observability and auditability that was previously impossible.

Data Fabric: the unified data backbone

Every orchestration is only as intelligent as the data it acts on. Even the most elegantly designed workflow can’t deliver consistent results if it’s fed by incomplete, outdated, or siloed information. In most enterprises, this remains the norm: customer records in CRM, transactions in ERP, risk data in spreadsheets, and operational metrics scattered across analytics tools.

Now in preview, UiPath Data Fabric closes this gap by serving as the unified data backbone for agentic automation. It connects live, trusted data from Snowflake, Databricks, and other systems directly into the orchestration layer using a zero-copy architecture (no data duplication), enabling every automation, app, and agent orchestrated in Maestro to operate with the same real-time truth.

These collaborations underscore our open philosophy: UiPath meets you where your data, systems, and AI investments already are, and connects them all under one governed platform for agentic automation and orchestration.