Automation Anywhere has announced a new AI + Automation Enterprise System infused with its second-generation GenAI Process Models to speed up discovery, development and deployment of AI process automations.

The company has also launched new AI Agents to manage complex cognitive tasks and automate more than ever before possible across every system in an enterprise.

A new feature is the ability to build custom AI Agents with the new AI Agent Studio. AI Agents have the ability to learn from enterprise data, make informed decisions, and take action responsibly across any enterprise system, speeding processes by up to 90 percent.

AI Agent Studio features low-code tools, making it easy for developers of all skill levels to quickly create specialized AI Agents to help with their specific use cases – no data scientist required. These AI Agents are adaptive, capable of learning from complex enterprise data, and able to take swift action for quick resolution and higher ROI.

Enterprise AI Agent features include:

New AI Agent Studio:, provides developers of all levels with low-code tools to easily build, manage, and govern custom AI Agents. Developers can start with the foundational model of choice, including models from AWS, Google Cloud, and Microsoft Azure OpenAI Service and more. Developers also will be able to augment AI Agents with enterprise knowledge through a native Retrieval-Augmented Generation (RAG) service, and Amazon Bedrock in October. Developers will have built-in prompt testing to ensure outputs are relevant for any use case before putting AI Agents into action.

Enhanced Security & Governance: equips AI Agents with built-in controls for the protection and monitoring of how AI and company data is being used to ensure security and compliance. New functionality includes monitoring and audit of agent and model performance, guardrails for consistent use, and human-validation and prompt testing to limit hallucinations and maximize quality of outputs. Future Data Masking will automatically redact sensitive data from any input getting processed by a model.

The underlying engines that power the AI + Automation Enterprise System are Automation Anywhere's GenAI Process Models. The GenAI Process Models 2.0 are designed to drive faster process discovery, 30 percent faster automation creation, 90 percent accuracy with document processing, and 50 percent more automation resiliency – above and beyond what LLMs alone can deliver.

The models are tuned with rich metadata from more than 300 million process automations running on Automation Anywhere's cloud-native platform.

A new set of developer automation solutions built on the new GenAI Process Models 2.0 include:

New Automator AI, generally available, a comprehensive set of generative AI products and capabilities to accelerate the automation lifecycle, making it faster and easier to build, deploy and manage automations. Features include:

- Generative Recorder, enabling teams to build more resilient UI automations with a generative AI fallback that automatically detects changes in a source application interface and self-heals in real-time to keep work moving, reducing automation downtime by up to 50 percent.

- Enhanced Autopilot, generally available, enables cross-functional teams to go from discovery to automation in record time by quickly converting process documentation into draft process automation using generative AI. Autopilot now allows for inputs from any mining tool in BPMN format to build automations.

Document Automation: leverages generative AI enhancements for realtime processing of any document type, including unstructured documents and achieves more than 90 percent accuracy.

Companies now can rapidly capture data in the flow of work from any document type with the new ability to extract data from complex tables, more than 30 supported languages, and expanded model options. Model setup and deployment is faster and easier with a new testing and setup experience and new support for on-premise deployment.

https://www.automationanywhere.com/