Smart Agent Is Here: Retrieval That Plans, Verifies, and Reaches Beyond Your Knowledge

Smart Agent adds live business data through MCP and web searches
by Michael Marolda Posted on October 02, 2026

Progress Agentic RAG's latest release introduces Smart Agent, an agentic retrieval capability that turns your knowledge layer into something that can reason across sources instead of just returning the nearest match. Rather than treating every question as a single lookup, Smart Agent breaks complex questions into sub-questions, decides where each one is best answered, checks whether what it found is actually enough, and keeps going until it is. It's available now as part of the same platform your teams already have configured — no new pipeline, no re-indexing.

Two things make this release worth paying attention to beyond the core capability: Smart Agent can now reach into the business applications your teams already run via MCP, and it ships with a built-in web search tool for pulling in information that lives outside anything you've indexed.

Querying the systems you already run, not just the ones you've indexed

Most of what an organization knows doesn't live in a document. It lives in a CRM, a support desk, a ticketing system — structured records that change by the hour. The Progress Agentic RAG Smart Agent connects to these through MCP, so it can pull live data from the same business applications your teams already depend on and combine it with whatever's sitting in your Knowledge Boxes.

A few ways that plays out in practice:

A support team using Zendesk could let Smart Agent answer "has this customer reported this issue before, and what resolved it?" by pulling open and historical tickets from Zendesk and cross-referencing them against product documentation and known-issue write-ups in the knowledge layer turning a search that used to require jumping between two systems into a single answer with both sources cited.

A sales or account team running Salesforce could ask a question like "what's the status of this renewal and has legal flagged anything?" and get back an answer that combines live opportunity and account data pulled from Salesforce with unstructured context — contract language, prior correspondence, meeting notes — that never made it into any CRM field to begin with.

A field service or IT operations team connected to a system like ServiceNow could have Smart Agent correlate an open incident with equipment history, prior resolutions, and a technical manual, producing a diagnosis with its sources attached instead of routing the question to whichever person happens to remember a similar case.

The pattern underneath all of these is the same: Smart Agent doesn't require you to have already pulled everything into your knowledge layer. It reaches out to the system of record at query time, through MCP servers you've likely already deployed for other purposes, so the connectors your team built keep paying off.

Reaching past your own walls with the new web search tool

Not every answer lives inside an organization's own content, and until now, extending Progress Agentic RAG's retrieval to the open web meant stepping outside the platform entirely. Smart Agent now includes a built-in web search tool — currently backed by Perplexity and Google Gemini — that can call as one more source alongside your Knowledge Box and connected business systems, and weigh in the same evaluate-and-iterate loop as everything else.

Practically, that opens up answers that combine internal and external context in a single response: a product team could ask a question that requires both an internal spec and current information about a competing product's public pricing; a compliance team could check an internal policy document against a regulation that changed last month; a research group could combine an internal report with a recent external finding it never had reason to index. In each case, Smart Agent decides for itself when the answer needs to leave your own content and reach the open web and still comes back with citations either way.

Together, these two additions are what make Smart Agent something closer to a general reasoning layer over everything your organization touches, rather than a smarter way to search documents you already indexed.

Learn more about the Smart Agent in Progress Agentic RAG today.


Author Michael Marolda in front of Niagara Falls
Michael Marolda

Product Marketing Manager, Senior

Michael Marolda is a seasoned product marketer with deep expertise in data, analytics and AI-driven solutions. He is currently the lead product marketer for the Progress Agentic RAG solution. Previously, he held product marketing roles at Qlik, Starburst Data and Tellius, where he helped craft compelling narratives across analytics, data management and business intelligence product areas. Michael specializes in translating advanced technology concepts into clear, practical business terms, such as Large Language Models (LLMs), Retrieval-Augmented Generation (RAG) and modern data platforms.
More from the author

Related Products:

Agentic RAG

Progress Agentic RAG transforms scattered documents, video, and other files into trusted, verifiable answers accelerating AI adoption, reducing hallucinations, and improving AI-driven outcomes.

Get in Touch

Related Tags

Related Articles

How Progress Agentic RAG Reached #1 on the Snowflake MADQA Leaderboard
Progress Agentic RAG ranked #1 on Snowflake's MADQA Leaderboard, an independent benchmark built with Oxford, UNC-Chapel Hill, and Hugging Face that tests whether AI answers are accurate and properly grounded in real documents, not just plausible-sounding. The win validates a year of work since the Nuclia acquisition building the Agentic Knowledge Layer (agentic capability embedded across ingestion, retrieval, and evaluation) and lands as the broader market shifts toward trust and governance as the deciding factor for enterprise AI.
Why AI Costs Spike After the First Use Cases
In this blog, we take a look at why AI costs accelerate so quickly after initial implementation success and how a modular approach to Agentic RAG can transform isolated pilots into a scalable, sustainable foundation for enterprise AI.
Progress Agentic RAG Adds a Smart Agent Plus Native Teams and WordPress Support
The October release adds three capabilities to Progress Agentic RAG with nothing to install: a Smart Agent that answers questions using both indexed content and live business apps like a CRM, a native Microsoft Teams app, and a WordPress plugin that keeps content current. Built-in citations, quality scoring and audit logs let reviewers trace each step the agent takes. The blog ends by inviting readers to an October 22 walkthrough or a demo.

Michael Marolda October 01, 2026
Prefooter Dots
Subscribe Icon

Latest Stories in Your Inbox

Subscribe to get all the news, info and tutorials you need to build better business apps and sites

Loading animation