The Future of Agentic Orchestration

v.38 of our Agentic Process Automation System is available today

Automation Anywhere is continuously developing Agentic Process Automation innovations that enable enterprises to automate up to 80% of their business processes. For our customers, today’s automation technologies provide a valuable platform for exceeding those expectations as technologies mature, needs evolve, and innovations appear — all of which is happening at an extremely rapid pace. What was visionary yesterday is now table stakes, and we never stop advancing our solutions.

We’re excited that Automation Anywhere was recognized in the first-ever Gartner® Magic Quadrant™ for Business Orchestration and Automation Technologies. We're also proud to be one of only 20 companies evaluated, given the hundreds of companies working in agentic AI. BOAT reflects the features available in v.37 of Automation 360, part of our Agentic Process Automation (APA) System. However, v.38 makes significant advances in our vision for autonomous AI agents that act with purpose and deliver groundbreaking new capabilities for agentic orchestration, extending the value of full-featured, multi-technology platforms.

The new capabilities in v.38 raise the bar significantly as we help enterprises modernize operations faster, safer, and at scale. This release also marks a fundamental shift from rules-based automation to goal-driven agents via Automation Anywhere’s comprehensive business orchestration platform, Mozart Orchestrator — all powered by the industry’s first Process Reasoning Engine (PRE). Among many other enhancements, our v.38 release delivers:

• Multi-agent orchestration with interoperability via the Model Context Protocol (MCP) and Agent-to-Agent (A2A) integrations
• Accelerated process developer productivity to streamline agentic process automation development, boost accuracy, and expand automation coverage
• Enhanced scalability and resiliency for the modern enterprise to run agents on public and private clouds with high-volume document processing and interruption-proof automations

Let’s explore what else we’ve packed into v.38 to help enterprises get on the BOAT — and extend it with agentic orchestration.

Evolving from legacy orchestration to agentic orchestration

Legacy orchestration platforms were architected around structured business processes that could be mapped, modeled, and managed. But the inherently rigid, rules-based logic limits this approach to very well-defined and static processes. When the process changes or a scenario falls outside the rules, humans are required to step in to fix it manually. The rise of cloud-native infrastructures and AI-driven solutions adds additional complexity as processes flow across tools, data, and systems that legacy orchestration platforms can’t handle effectively. The results are quickly outdated process maps, growing development backlogs, and deteriorating operational agility.

The modern approach to orchestration focuses on goals rather than tasks. We now have AI agents that understand context, reason through problems, plan actions, and adapt to shifting scenarios as they achieve goals. APA puts these agents to work, enabling them to develop a fluid path, dynamically adapt to changing contexts, variables, and exceptions, and reason around them without human intervention. Better yet, APA coordinates and governs multiple AI agents, RPA bots, APIs, and humans as they collaborate across complex workflows and enterprise ecosystems in pursuit of their ultimate goals.

The result is APA that’s not just an improvement over legacy orchestration — it's an entirely different approach to enterprise automation.

The Agentic Advantage

A bold step forward for agentic automation with v.38

We keep pushing beyond our customers’ expectations to bring agentic automation to their most mission-critical enterprise operations. Our solutions prove that AI agents can think, act, and work together across systems, tools, and platforms at scale, with governance, and in production with our APA System.

From banking reconciliation cycles to healthcare claims processing to global supply chain optimization, Automation 360 v.38 is built for high-stakes operations where speed, security, and reliability matter most.

Below, we’ll cover several key advancements in our latest release and show how they’ll impact your automation success. For a deeper look at these new capabilities, check out our live webinar on Oct. 28, or register to attend our upcoming virtual Pathfinder Summit on Nov. 18-19. And, as always, more details are in the v.38 release notes.

Autonomous AI agents that act with purpose

Autonomous AI agents enable a leap from rule-based automation to goal-driven agents powered by PRE. These agents adapt in real time, orchestrate tasks across bots, APIs, and other agents, and embed human-in-the-loop checkpoints, all with auditability and governance built in.

For the first time, enterprises can automate workflows that demand judgment and adaptability, unlocking entire classes of processes previously beyond the reach of automation. Key capabilities include:

• Real-time adaptation as agents adjust intelligently to changes using structured prompts and contextual variables
• Tool-calling at scale to leverage existing automations, APIs, and other AI agents, along with tools from third-party applications and enterprise systems
• Human-in-the-loop controls that add checkpoints for approvals, escalations, and exception handling across complex processes
• Enterprise-grade governance to equip developers with detailed execution logs, testing tools, and override controls for reliable deployment

In v.38, we’re also providing trusted, ready-to-use generative AI models to help solve one of the biggest challenges in enterprise AI adoption: the time-consuming process of obtaining model approvals. It’s typically those lengthy InfoSec reviews that slow down or even block AI deployments. We’re eliminating this barrier with pre-approved, hosted models directly within A360. Here’s where you’ll see the benefits:

• Instant adoption by using trusted models on day one and with no additional integrations required
• Enterprise-grade compliance because all models undergo rigorous internal security and governance reviews
• Secure runtime with models run within Automation Anywhere–managed cloud environments
• Seamless integration to apply models immediately to AI Skills and AI Agent reasoning steps

Note that provided models are available exclusively to public cloud AWS customers. Google Cloud is expected to be supported later this year.

To help AI agents act with purpose across tools and solutions, v.38 introduces our Mozart Orchestrator and expands the capability for multi-agent enterprise ecosystems with MCP and A2A interoperability. The new inbound MCP support, powered by PRE, enables intelligent, user-aware, and multi-tenant collaboration between APA automations and any MCP-compliant agent. PRE adds a reasoning layer that understands automations, generates metadata, and maps intent for governed, compliant execution. Cross-agent collaboration through A2A lets APA AI agents exchange data, delegate tasks, and coordinate workflows with third-party agents for use cases, such as planning, logistics, and analytics. And, our open, standards-based architecture prevents lock-in and ensures seamless integration with internal or partner-built agents.

How does this come to life? Consider a manufacturer’s APA supply planning agent that collaborates with a third-party logistics partner’s agents via A2A, synchronizing production schedules with shipment updates in real-time to reduce delays and ensure on-time deliveries.

Note that MCP and A2A support are released as a developer preview. Please contact your Automation Anywhere team to enable this feature.

Design and governance in one

In v.38, Process Composer provides a single environment for enterprises to design and govern processes across agents, bots, APIs, and documents. Key capabilities include:

• Multi-agent orchestration to design workflows up to 60% faster and that enable multiple agents, bots, and APIs collaborate seamlessly to drive mission-critical operations
• Contextual next-step suggestions with PRE-powered recommendations that guide builders with intelligent next actions to reduce design complexity and cut up to 70% from build time
• Drag-and-drop simplicity enables teams to orchestrate AI agents within a single Process Composer canvas just like bots and APIs, and as opposed to hopping between tools

Insurance firms can put this into action for claims processing, for example, which can be accelerated by orchestrating agents for eligibility verification, policy validation, and payment approvals. This enables a single governed workflow that can shorten cycle times by up to 50%.

Build faster and deploy with confidence

Speaking of Process Composer, new features in v.38 streamline development, boost accuracy, and expand automation coverage. That helps you scale APA solutions by eliminating manual setup, testing hurdles, and integration constraints.

Co-Pilot for Automators in Process Composer helps developers build workflows up to three times faster using natural language prompts powered by PRE, with best practices (naming conventions, variable setup, and exception handling) applied automatically. End-to-end queue testing in Editor adds validations for complete workload automations directly in the build environment, simulating live queues and ensuring business logic works as intended.

These enhancements bring enterprise-grade productivity to both professional developers and citizen builders so teams can deliver faster, safer, and more precise automations that scale to mission-critical use cases.

Elastic runtime engine for AI agents, now on private clouds

Enterprises increasingly require cloud-scale automation while maintaining complete control over their environments. With v.38, we’re delivering the agentic runtime engine directly into private cloud deployments.

Here’s what you can expect:

• Cloud-native inside your firewall, so you can deploy the same proven runtime used in public cloud, now within your private infrastructure
• Elastic auto-scaling to seamlessly scale API and AI agent workloads up or down based on real-time demand
• Strengthened compliance that keeps sensitive data within customer-controlled infrastructure, ideal for highly regulated industries such as finance, healthcare, and government
• Performance at enterprise scale with local execution that reduces latency for real-time use cases, such as fraud detection, order fulfillment, and transaction monitoring

We’re the only vendor delivering a unified APA engine across both public cloud and private VPCs, enabling actual hybrid operations without compromise.

Note that the runtime engine is currently available for private cloud customers on AWS and all public cloud customers.

Document processing without limits

Enterprises that depend on document-heavy workflows often struggle with manual review bottlenecks and local infrastructure constraints. With v.38, we’re introducing a Cloud Extraction Service that delivers high-volume document processing through a secure, distributed cloud infrastructure.

Cloud Extraction Service provides:

• Massive scale to efficiently process large volumes of documents, such as claims, invoices, and contracts, with elastic capacity that scales with demand
• Adaptive generative AI queries that automatically adjust to layout variability across thousands of formats and dynamically apply the correct query at run time why
• Infrastructure-free to offload processing from local bot runners and VMs to our secure cloud endpoint

A global insurer, for example, can now process claims at scale without adding infrastructure or managing dozens of bot runners — achieving 5x faster document processing speeds while cutting costs and operational overhead.

Note that Cloud Extraction Service is available exclusively to public cloud customers.

Continuing to enhance enterprise resiliency

Enterprises require automation that is reliable, resilient, and capable of extending into complex, legacy, and virtualized systems. With v.38, we’re enhancing resiliency-focused innovations to ensure APA automations remain robust, interruption-proof, and enterprise-ready.

Key capabilities include:

• Interrupt handler for built-in automation continuity
  • Detects unexpected pop-ups or re-authentication prompts (e.g., session timeouts, confirmation dialogs) in real time
  • Automatically executes configured resolutions, such as clicking "OK" or re-login sequences
  • Resumes workflows seamlessly without manual intervention
• Structured data extraction (limited availability)
  • Easily capture structured, pattern-based data (tables, lists, repeating fields) from web applications
  • Simplifies what were once complex automations, improving accuracy and boosting developer productivity
• Expanded Citrix automation average
  • Object-based automation now works with Citrix apps even when window titles are inconsistent or dynamic
  • Reduces dependency on fragile image-based or keystroke automations, delivering greater stability for legacy system use cases

We never stop innovating

With v.38, we’re setting a new standard for innovation in automation by empowering enterprises to modernize mission-critical operations faster, more securely, and at scale. We’re excited for you to experience these latest advancements across your operations.

To learn more, register for our live product webinar on October 28.

Gartner Magic Quadrant for Business Orchestration and Automation Technologies, Saikat Ray I Tushar Srivastava I Marc Kerremans I Arthur Villa I Cathy Tornbohm I Sachin Joshi October 15, 2025. Gartner does not endorse any vendor, product or service depicted in its research publications and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner’s Research & Advisory organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose. GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally, Magic Quadrant is a registered trademark of Gartner, Inc. and/or its affiliates and is used herein with permission. All rights reserved.