This Is How the Autonomous Enterprise Becomes Real

Today at Imagine 2025, we didn’t just launch features. We revealed a new blueprint for work. With the Agentic Process Automation (APA) System, enterprises can finally turn ideas into action—instantly and confidently.

Powered by the industry-first Process Reasoning Engine (PRE), and expanded with Enterprise UI Agents, agentic orchestration, and conversational human-agent collaboration, the APA System marks the beginning of agentic enterprise operations at scale. It doesn’t just automate tasks—it orchestrates outcomes across AI agents, automation, and people, reshaping the way work happens.

“With the APA System, we’ve built more than a product—we’ve engineered a new way of working,” said Adi Kuruganti, Chief Product Officer at Automation Anywhere.

“Powered by our Process Reasoning Engine, AI agents can now understand goals, interpret context, and orchestrate work across AI agents, traditional automation, and people in real time. Enterprises can now safely expand much further in what can be automated for mission-critical operations.”

Here's what we unveiled on the Imagine stage:

The process intelligence core: Process Reasoning Engine (PRE)

Now GA – Powers the Agentic Process Automation (APA) System

PRE is unique to Automation Anywhere—a first-of-its-kind AI automation model designed to drive agentic automation for complex enterprise operations.

Powers generative automation and self-healing workflows
Extracts structured data with 90%+ accuracy from any document
Delivers 3x higher automation efficacy and 60% greater resiliency
Supports emerging standards like GCP’s Agent-to-Agent (A2A) protocol and Anthropic’s Model Context Protocol (MCP)
Enables AI agents to plan, act, and learn continuously

PRE moves automation from intention to action. It’s how your enterprise thinks, adapts, and scales—like never before.

Enterprise UI Agents: Automate any interface like a human

Now in Preview

One of the biggest breakthroughs at Imagine 2025 is the launch of adaptive “computer use” AI agents that mimic human behavior to automatically navigate across highly variable web interfaces.

These UI Agents can:

Navigate web apps with human-like understanding
Adjust dynamically as layouts and interfaces change
Handle multi-step, complex workflows like real-time credentialing, identity verification, competitive pricing updates
Unlock long-tail use cases previously considered too complex

For companies like Alight, which support over 1,000 organizations, including a majority of the Fortune 100, adaptive Enterprise UI Agents deliver transformative impact. During open enrollment, they manage thousands of unique portals and validation rules for over 35 million people. This process used to take months of manual work and large teams to complete.

Now, with UI Agents, that work is automated end-to-end. A simple prompt kicks off the process, and the agents handle login, navigation, and rule validation just like a human, delivering results in days instead of months.

Agentic orchestration: Automating mission-critical work across AI agents, automation, and people

Now GA | Parallel Processing: Preview Q2 2025

Modern operations are messy, complex, and dynamic. With our APA system’s powerful orchestration engine, you can now coordinate AI agents, RPA bots, APIs, documents, and people in real time, all in one unified, enterprise-grade system.

Here’s what makes the unique orchestration engine work:

Model business logic with BPMN 2.0 drag-and-drop workflow design for faster scaling (NEW)
Orchestrate AI agents, bots, APIs, documents, and human experts in a single flow (NEW)
Monitor AI agents actions with real-time analytics and observability (NEW)
Automate with reliability using event-driven triggers and upcoming support for parallel processing (NEW)
Seamlessly orchestrate across both internal and external agents, including those built on leading platforms such as AWS Bedrock, Google Agentspace, Microsoft CoPilot, and Salesforce Agentforce, as well as developer frameworks such as Langchain and CrewAI (NEW)

By combining cognitive AI agents with deterministic automation on a single enterprise-grade platform, the APA System enables organizations to drive mission-critical work with greater flexibility and reliability.

Build, deploy & govern process AI agents at scale

Now GA | Agent Evaluation & Prebuilt AI Agent Library: Preview H2 2025

As enterprises adopt more AI agents and scale agentic automation, the challenge shifts from “can we build it?” to “can we manage it at scale with confidence?”

The enhanced AI Agent Studio answers that call. It brings discipline, speed, and control to the full AI agent lifecycle.

Build using any GenAI model, framework, or tool—OpenAI, Anthropic, LangChain, CrewAI (GA)
Deploy flexibly across cloud and private environments (GA)
Protect data with built-in PII masking, monitoring, and audit (NEW)
Fast-track results with a growing prebuilt agent library (H2)
Validate and score agent readiness with our new evaluation service (H2)

Together, these enhancements bring enterprise-grade governance to AI agent deployment—so you can move faster, stay secure, and keep full visibility as your agents scale.

Real-time observability: Enterprise control and confidence

Now GA | Agent ROI Tracking: Preview H2 2025

One of the biggest blockers to scaling agentic automation has been visibility. Leaders often ask: What’s actually running? Is it working? What’s the impact?

With Automation Anywhere’s APA System, the new observability dashboard provides a real-time command center for managing automation across AI agents, automation, and people:

Get a unified view across models, AI agents, and processes to monitor performance and system health (GA)
Reduce operational costs by identifying bottlenecks in throughput, latency, and success rates across agents and systems (GA)
Troubleshoot live processes and agent interactions with real-time diagnostics and failure tracing (NEW)
Track agent usage, updates, and outcomes over time with full lifecycle visibility and business impact measurement (H2)
Monitor how multiple agents work together across distributed workflows to optimize coordination and efficiency (H2)

With these built-in security and governance capabilities, the APA system enables organizations to scale automation with clarity and confidence.

Human-agent collaboration: Enterprise automation becomes conversational

Now GA | Multimodal Interface: Preview H2 2025

For the first time, enterprise automation becomes truly conversational. With the APA System, users can now interact with process AI agents using natural language, and get back real results, grounded in enterprise data.

Turn intent into action in any app, without clicks or training (GA)
Get enterprise-specific answers—not generic LLM responses—grounded in your company’s policies, data, and workflows (GA)
Personalize agent behavior based on user role, recent activity, and system permissions (H2)
Communicate more easily via voice and video in an upcoming multimodal interface (H2)

When agents understand your intent and respond with real outcomes, instantly and accurately, they become true collaborators, unlocking a smarter, faster way to get work done.

Process any document with Document Automation

Now GA | MS Embedded Intelligence & Automated GenAI Learning: Preview H2 2025

From claims processing to vendor onboarding, documents are still one of the most persistent sources of operational friction. Legacy OCR and rule-based tools break easily, require heavy maintenance, and don’t scale to meet today’s complexity.

With GenAI-powered understanding and extraction, Document Automation turns the unstructured data into accurate, actionable insights, dramatically expanding what enterprise workflows can achieve.

Extract from any format or layout with 90%+ accuracy (GA)
Benchmark and optimize quickly with built-in testing (NEW)
Process millions of documents per month with audit-ready scalability (NEW)
Tap into Microsoft Document Intelligence 4.0 for deeper classification (H2)
Continuously improve with self-learning document models (H2)

The 3x YoY growth in page volume processed by our customers shows it works—at scale.

Accelerate time to value with pre-built agentic solutions

The types of critical work we can transform with process AI agents are virtually limitless. But we’re starting with where the need is urgent and the impact is immediate.

Leading the way is our fully packaged accounts payable solution, now generally available. Built on the APA System, it combines agentic automation—pre-trained AI agents with deep domain knowledge—and the Agentic Workspace, a modern interface that brings tasks, data, and decisions directly to the user. Together, they automate complex invoice processing workflows from intake to exception handling, delivering outcomes fast.

And that’s just the beginning.

We’re expanding the portfolio with pre-built solutions for customer support, KYC in banking, and revenue cycle management (RCM) in healthcare—each designed to handle complex, cross-functional work with speed, accuracy, and enterprise-grade integration.

We’re also developing specialty agents for tariff classification, competitive intelligence, and employee experience, with targeted availability in the next 1–2 quarters.

These solutions are built on the APA System, turning business users’ intent into action through natural conversation—getting work done where it matters most.

If you’re wondering where to begin, start here.

This is how organizations move from potential to value—fast.

Ready to see it in action?

Connect with your Automation Anywhere team to see how the APA System can transform your operations—and take the first step toward your autonomous enterprise.

Let’s get started.