Enhanced knowledge processing, voice capabilities, and platform performance to improve agent quality, usability, and responsiveness.

1. KnowledgeHub Enhancements: Added Xiaohongshu OCR and proprietary OCR support to improve image and document parsing.
2. KnowledgeHub Web Sync: Integrated crawler capabilities and FireCrawl for more efficient web content synchronization.
3. Voice Enhancements: Added voice preview to help users evaluate and select voice options more easily.
4. Self-Play Enhancements: Support multilingual test script generation to improve testing across global use cases.
5. Performance Improvements: Optimized page loading speed and overall user experience for a faster, smoother workflow.

Enhanced operational insights, agent configuration, and runtime experience to improve usability, real-time performance, and operational efficiency.

1. Insight Enhancements: Added an operations dashboard to help monitor agent performance and business metrics more effectively.
2. Observation Configuration: Introduced Observation configuration for more flexible runtime analysis and business rule customization.
3. Lite Mode Expansion: Extended Lite Mode support to Preview and Deploy, bringing low-latency performance to more agent workflows.
4. Build Improvements: Simplified the Build experience, improved error handling, and added a sign-out option on the initial setup page for easier account switching.

Enhanced agent orchestration, knowledge management, and continuous optimization to build more intelligent and maintainable enterprise AI agents.

1. Build Enhancements: Added support for Relationships and Retrievers, enabling more advanced workflow orchestration and knowledge retrieval.
2. Suggestions: Introduced the Suggestions module to identify optimization opportunities and continuously improve agent performance.
3. KnowledgeHub Enhancements: Added QA import and query-based knowledge matching to improve knowledge management and answer accuracy.
4. Self-Play Evaluate Improvements: Refined evaluation fields and result presentation to make test analysis and agent optimization more efficient.

Enhanced agent configuration, testing, and knowledge management to improve development efficiency, quality, and business adaptability.

1. Greeting Configuration: Enable or disable greetings during deployment and support dynamic variables for more personalized conversations.
2. Import & Export: Import and export Build configurations or individual journeys to simplify reuse and cross-environment migration.
3. Voice Preview: Test voice interactions before deployment to validate user experience and improve voice agent quality.
4. Test Enhancements: Added Self-Play and Evaluate capabilities to automate simulation and evaluation, accelerating agent optimization.
5. KnowledgeHub Enhancements: Support QA import, intelligent parsing, and knowledge retrieval to improve knowledge management efficiency and response accuracy.

Enhanced agent integration, deployment, and openness capabilities, enabling broader business automation and voice-based use cases.

1. Tool Configuration: Build now supports HTTP tool creation and invocation, allowing agents to connect with external systems and automate business workflows.
2. Voice Channel Deployment: Deploy now supports voice channels with automatic phone number provisioning and voice agent configuration, accelerating voice agent rollout.
3. Chat OpenAPI: Introduced Chat OpenAPI, enabling developers to access agent conversation capabilities through standard APIs and integrate them into business systems and applications.

Expanded orchestration, configuration, and evaluation capabilities to enable more sophisticated and scalable AI agent deployments.

1. Flexible Agent Creation: Create agents from scratch or start with templates to balance flexibility and speed.
2. Multi-Journey Support: Manage multiple workflows within a single agent, reducing maintenance effort while supporting more complex use cases.
3. Enhanced Configuration: Support variable write-backs and advanced workflow controls for greater business flexibility and dynamic decision-making.
4. Self-Play Evaluation: Automatically evaluate agent performance through simulated conversations, helping teams identify configuration issues and workflow risks before deployment.

Enhanced agent testing, runtime performance, and platform governance to help teams identify issues faster and deliver more reliable AI experiences.

1. Self-Play Simulation: Automatically generate conversation scripts to reduce manual testing effort and improve scenario coverage.
2. Lite Mode: Run agents in Lite Mode to significantly reduce latency and deliver a smoother experience for voice and other real-time interactions.
3. Built-in Counter Tool: Support complex state management and conditional logic, expanding agent capabilities in rule-driven workflows.
4. Data Access Controls: Strengthened data isolation and security to support multi-team collaboration and enterprise governance.

Delivered the core capabilities required to create, test, monitor, and deploy AI agents, enabling business teams to build and validate agents faster and bring them into production with confidence.

1. Login & Registration: Established a unified account system, making it easier for users to get started with building and managing agents.
2. Agent Creation: Automatically generate onboarding flows to reduce setup effort and accelerate time to launch.
3. Knowledge Management: Centralize and maintain knowledge to improve response accuracy and consistency.
4. Preview Testing: Test conversations with agents in real time and view matched rules and execution paths, making it easier to validate configurations and optimize agent performance before deployment.
5. Batch Testing: Validate agent performance across multiple scenarios before launch, reducing production risks.
6. Conversation Monitoring: Access detailed conversation logs to quickly identify issues and continuously optimize agent performance.
7. Web Deployment: Publish agents to web channels with one click, accelerating business adoption and validation.