What's new
This release of HCL UnO Agentic AI Builder introduces significant enhancements designed to increase configuration flexibility, strengthen model management, and improve the overall user experience. These updates are organized into three key areas:
- Build Smarter, More Capable Agents
-
Create Agents for Complex Tasks
Build agents capable of handling interconnected workflows with higher reliability and efficiency.
-
Choose the Best Model for Your Agent
Select models from leading providers—AWS Bedrock, Google Vertex, Microsoft Azure, NVIDIA, and OpenAI—or run inference locally using Ollama-deployed models within your data center.
-
Seamless Agent-to-Agent (A2A) Integration
Design sophisticated multi-agent solutions by integrating A2A agents with platform-native agents for advanced collaboration.
-
Advanced Task Orchestration
Implement conditional hand-offs and decision points for more accurate, reliable, and cost-efficient executions.
-
Enhanced Testing & QA
Validate real-world scenarios with an enriched testing interface featuring new KPIs, detailed tracing, and support for PDF attachments.
-
- Accelerate Time-to-Value
Build and deploy high-quality agents faster than ever with streamlined processes and guided workflows.
-
Streamlined Creation
Manage all LLM/SLM models from a new dedicated profile page and set default attributes to simplify initial agent setup.
-
Guided Template Transformation
Receive real-time guidance when converting templates into fully functioning agents.
-
Simplified Installation
Enjoy a smoother setup process with automatic deployment of prerequisites for QA and staging environments.
-
- Boost Platform Performance and Efficiency
Experience a faster, more responsive platform built to scale.
-
Scalable Agent Execution
Handle growing workloads with horizontal scaling support, ensuring consistent agent performance as usage expands.
-
Improved User Experience
Enjoy a more responsive interface with lazy loading, collapsible sections, enhanced search tools, and a dedicated page for managing model preferences.
-