Named in Gartner's Market Guide for Agentic Analytics, two years running.
Gartner published its second Market Guide for Agentic Analytics in February 2026. Unsupervised was named a Representative Vendor in both editions. The report's most notable new prediction: by 2028, 60% of agentic analytics projects relying solely on MCP will fail because they lack a consistent semantic layer.
February 13, 2026
The prediction
“By 2028, 60% of agentic analytics projects relying solely on MCP will fail due to the lack of a consistent semantic layer.”
Gartner, "Market Guide for Agentic Analytics," Deepak Seth, Georgia O’Callaghan, Fay Fei, Jeroen Cornelissen, 9 February 2026.
That prediction explains why so many AI-for-analytics pilots stall after the initial demo. An LLM can generate a SQL query. But “revenue” means different things in different tables, “churn” is calculated differently by different teams, and no amount of prompt engineering fixes that. Gartner calls semantic alignment “foundational for effective agentic analytics” in both the 2025 and 2026 editions. In 2026, they put a number on it.
One year on
How the market grew in one year.
The 2025 edition was Gartner's first Market Guide for this category. It described a market of about 30 vendors, mostly in pilot and experimentation mode. One year later, the 2026 edition counts 37 vendors and describes a market entering early-scale deployment. Established ABI vendors (Databricks, Snowflake, SAP) are adding agentic features. New specialized players are emerging alongside them.
The report cites adoption data from the 2025 Gartner Generative and Agentic AI in Enterprise Applications Survey.
Have deployed or are actively deploying AI agents
Have deployed or are exploring goal-driven agents that operate autonomously
Of IT leaders expect significant productivity impact
Source: Gartner, “Market Guide for Agentic Analytics,” 9 February 2026.
Three ideas
Three ideas in the report worth paying attention to.
1 · “Perceptive analytics” is a new concept
The 2026 edition introduces a term that wasn't in the 2025 report: perceptive analytics. The idea is systems that don't wait for a query. They continuously monitor data and surface context-aware insights on their own, detecting anomalies, spotting emerging trends, and flagging risks before anyone asks. Gartner argues that organizations adopting this approach gain “real-time situational awareness that enables proactive, context-aware decisions rather than reactive analysis.”
This draws a line between two kinds of agentic analytics: reactive (answer questions faster) and proactive (find things no one thought to look for). Most current tools are reactive. The report suggests the value shifts toward proactive over time.
2 · LLMs and production analytics are diverging
“By 2028, 60% of users will use general-purpose LLMs for ad hoc and exploratory analysis, while production-grade reporting will remain in traditional ABI platforms.”
Gartner, "Market Guide for Agentic Analytics," Deepak Seth, Georgia O’Callaghan, Fay Fei, Jeroen Cornelissen, 9 February 2026.
In other words: ChatGPT is great for asking a quick question about your data. It's not what you run your weekly business reviews on. The report sees these as two separate markets with different requirements for accuracy, governance, and repeatability.
3 · Differentiation is moving from models toward governance
The 2026 report is explicit: market leaders will emerge based on “semantic consistency, explainability, cost controls, and delegation frameworks, rather than model sophistication alone.” The vendors that win won't be the ones with the best LLM. They'll be the ones whose agents produce reliable, auditable results an enterprise can trust at scale.
The report organizes vendors into three categories: traditional ABI platforms adding agentic features, capability-centric platforms (startups with deep technical specialization), and domain-specialized platforms (industry-specific). Differentiation is happening across all three.
Sources: Gartner, “Market Guide for Agentic Analytics,” Anirudh Ganeshan, Souparna Palit, David Pidsley, 21 February 2025; and Gartner, “Market Guide for Agentic Analytics,” Deepak Seth, Georgia O'Callaghan, Fay Fei, Jeroen Cornelissen, 9 February 2026.
Where we fit
The semantic layer problem is the reason Unsupervised exists. Our platform is built around a semantic model that maps business entities, relationships, and policies across an organization's data, so “revenue” means the same thing regardless of which database it comes from. Our agents don't generate SQL from LLMs. They operate on this structured layer.
We deploy three types of agents:
Discovery Agents
Run continuously against customer data, surfacing patterns and anomalies without a human asking a question. This is perceptive analytics in practice.
Conversational Agents
Answer questions in natural language, backed by structured semantic access rather than LLM-generated SQL.
Predictive Agents
Build and run models without code or infrastructure management.
These agents are in production at enterprise scale. Insights AT&T put into action carry an estimated value of more than $100M. A Fortune 500 healthcare payer found $58M. Across all deployments, Unsupervised has surfaced over $1B in actionable value.
“Unsupervised's AI Data Analysts have delivered strong ROI—improving our key metrics while empowering our team with faster, smarter access to data insights.”
Sources
Gartner, “Market Guide for Agentic Analytics,” Anirudh Ganeshan, Souparna Palit, David Pidsley, 21 February 2025.
Gartner, “Market Guide for Agentic Analytics,” Deepak Seth, Georgia O'Callaghan, Fay Fei, Jeroen Cornelissen, 9 February 2026.
GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved.
Gartner does not endorse any vendor, product or service depicted in its research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner's research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose.