Progress Software has released a new version of its Telerik and Kendo UI development tools, adding AI features designed to help developers build applications that work with both human users and AI agents. The Burlington, Massachusetts-based company announced the release on Aug. 20, positioning it as a way for engineering teams to accelerate UI development while preparing their software for agent-driven interactions.
The release embeds AI into several stages of the development workflow, from UI generation to debugging to document processing. It includes an Agentic UI Generator that bakes accessibility compliance into interface creation, and Fiddler Agent Skills that give coding assistants direct access to network traffic data for smarter debugging. The company also introduced WebMCP, a protocol designed to let AI agents interact directly with application interfaces instead of relying on fragile screen-scraping techniques.
"Engineering leaders need to both enable their teams to use AI to scale their productivity and to increase their production of agentic applications," said Loren Jarrett, EVP and GM, Digital Experience, at Progress Software. "This release enables engineering teams to meet those needs by extending developer productivity and providing pioneering UI tools to build AI-embedded applications compatible with both human and agent interactions."
Agent-ready features for modern development
Several new capabilities target the shift toward agent-driven software. Telerik Document Processing Libraries now include agent tools that can extract structured data from documents, analyze Excel files, convert formats, and generate new documents without requiring external AI services. Developers can also connect to Progress's RAG-as-a-Service product to add retrieval-augmented generation capabilities to .NET applications, letting large language models analyze unstructured enterprise data.
For teams maintaining older Windows applications, the release includes automated conversion to Telerik UI for WinForms using intelligent control mapping, code parsing, and MCP-based automation. New UI components such as Smart Paste, PromptBox, and SmartBox help developers build interfaces designed for AI interaction, while expanded framework support and faster CLI onboarding aim to shorten the path from project setup to working code.
Debugging with direct network visibility
Fiddler Agent Skills give coding assistants a view into network traffic, which means developers can receive proactive alerts about problems along with suggested solutions without leaving their integrated development environment. The feature addresses the common pain point of debugging AI-generated code, where errors often surface in network requests rather than in the code itself.
These updates arrive as engineering leaders face pressure to adopt AI while maintaining quality and governance. For developers, the practical takeaway is that AI-assisted coding is expanding beyond code completion into areas like accessibility compliance, document handling, and agent interaction - all of which are expected to become standard parts of building modern applications. As AI agents begin to use software directly, tools that support both human and machine interfaces will likely become a baseline requirement rather than a differentiator. For developers looking to build these skills, resources like an AI Learning Path for Software Developers can help bridge the gap between traditional UI development and agent-ready architecture. The broader trend in AI for IT & Development suggests that engineers who learn to work with agentic interfaces and AI-assisted tooling will be better positioned as these capabilities become standard.
Your membership also unlocks:
