Supper is an AI-powered business intelligence and analytics platform that helps people ask questions of company data in natural language. Instead of assigning every workflow to one general-purpose AI agent, Supper uses specialized agents for different modes of data work.
Supper uses multiple AI agents because data analysis, semantic model maintenance, and dashboard construction require different context, tools, and definitions of success. Supper's three major customer-facing agents are the Data Chat Agent, the Semantic Agent, and the Dashboard Builder Agent.
Why Supper Moved Beyond a Single AI Agent
The first version of Supper had one agent. Its job was straightforward to describe, even if it was technically difficult to deliver: give someone an AI agent that behaves like a data analyst. A user could request data, have Supper retrieve and manipulate it, explore what was available, generate a chart, and continue investigating through conversation.
That original agent became the Data Chat Agent, or DCA, and it remains at the center of the Supper experience. As the product expanded, however, an interesting problem appeared: a conversational interface can make every job look like the same job.
When people see an empty prompt box, they naturally assume they can ask it to do anything. Unlike traditional software, there may be no obvious settings page, dashboard editor, or modeling screen signaling that the user has entered a different mode of work. One response would have been to make the original agent responsible for everything, but we decided not to.
Working with business data involves several fundamentally different activities. Open-ended analysis requires an agent to retrieve data, inspect schemas, manipulate results, conduct statistical analysis, and create visualizations. Maintaining a semantic model requires an agent to understand business definitions and translate them into precise specifications, while dashboard construction requires decisions about datasets, charts, KPI tiles, filters, themes, and layouts.
Giving all of those responsibilities to one agent would mean constantly carrying tools, instructions, and context that are irrelevant to the user's current goal. More capability can also create more complexity, so Supper gives different agents different jobs.
What Are Supper's Specialized AI Agents?
Supper's three major customer-facing agents are:
- The Data Chat Agent, for conversational data analysis
- The Semantic Agent, for creating and refining business definitions
- The Dashboard Builder Agent, for turning data into reusable dashboards and visual components
The Data Chat Agent
The Data Chat Agent is Supper's general-purpose conversational analytics agent. Customers use it to ask questions about company data, retrieve relevant information, explore available schemas, manipulate results, perform statistical analysis, and create charts and visualizations.
The easiest analogy is a data analyst at your company. You do not need to know which tables contain the answer or exactly which transformations will be necessary. You describe the analytical goal, and the Data Chat Agent uses the available context and tools to work toward an answer.
Underneath the conversational experience, Supper gives the DCA access to capabilities that can include data retrieval, Python-based manipulation, schema exploration, statistical analysis, deeper inspection of returned datasets, and chart creation. That breadth is appropriate because general data analysis is itself a broad job.
The Semantic Agent
The Semantic Agent helps administrators create, inspect, and refine Supper's semantic model. A semantic model captures the business meaning an analytical system needs to interpret company data correctly, including business Terms, schema definitions, Skills, technical specifications, and descriptions of important concepts.
For example, a company may need precise definitions for an active customer, annual recurring revenue, customer churn, a qualified opportunity, or net revenue retention. These definitions can be cumbersome to build and maintain manually, and defining a metric is a different job from analyzing that metric.
The Semantic Agent is designed to work on the model itself. It can explore the model, investigate relevant schema details, refine descriptions, and assist with the specifications needed to make a business Term precise enough for reliable analytical work. Relevant Terms can then become part of the context Supper uses to interpret business concepts and retrieve data.
The Dashboard Builder Agent
The Dashboard Builder Agent helps users create and organize the datasets and visual components that make up a dashboard. Dashboard construction introduces a different mode of work because the question is no longer only, "What does the data say?"
Users must also decide which datasets belong together, whether a result should become a chart or KPI, which visualization communicates it most clearly, how filters should work, and how the dashboard should be organized. These decisions are related to data analysis, but they are not identical to it.
The Dashboard Builder Agent focuses on creating datasets, charts, KPI tiles, and other visual components, then helping organize them into a coherent dashboard. A useful result discovered through the Data Chat Agent can therefore become a reusable dashboard component without requiring the user to start over.
Why Not Give One AI Agent Every Tool?
If AI agents can use tools, why not give one agent every available tool? The answer is that tools are only half of the problem; intent matters too.
Consider three requests someone might make to a human data team:
- "Help me investigate customer churn."
- "Let's fix how our company defines an active customer."
- "Turn this analysis into an executive dashboard."
A talented person might be capable of participating in all three activities, but the mode of collaboration changes. The information they need, the definition of a successful result, and the appropriate next action are different in each case.
Supper's agent boundaries work the same way. The product does not need a separate agent for every small action, but specialization becomes useful when the customer's goal requires a different working context, set of tools, or definition of success.
The Benefits of Specialized AI Agents for Data Work
Specialized agents allow Supper to align the agent's working context with the task at hand. That creates several practical advantages:
- More relevant context: The agent can focus on the information needed for the current workflow.
- Clearer intent: The agent knows whether it is analyzing data, maintaining business definitions, or constructing a dashboard.
- Purpose-built tools: Each agent can prioritize the capabilities that support its specific job.
- Better workflow boundaries: Users can understand what kind of result the agent is trying to produce.
- Reduced complexity: A single conversation does not need to carry every possible instruction and responsibility at once.
The goal is not specialization for its own sake. It is to give users the shortest path to the right kind of work.
Making a Multi-Agent Analytics Product Feel Seamless
Users already understand boundaries in traditional software. They go to a settings page to change settings and open a dashboard editor to edit a dashboard because the surrounding interface tells them which mode they are in.
Chat changes that expectation. When users see a blank prompt, they naturally try to do everything there, so a multi-agent product must make its boundaries understandable without making the experience feel restrictive.
Supper approaches this in two ways. First, each agent needs to understand its purpose and direct users toward the appropriate workflow when a request falls outside that purpose. Second, the surrounding interface must make it clear whether the customer is performing general analysis, working on the semantic model, or building a dashboard.
The goal is not to prevent exploration. It is to help users reach the appropriate workflow without requiring them to understand Supper's underlying architecture.
The Future of Supper's Multi-Agent Architecture
The long-term challenge is making specialization feel seamless rather than fragmented. Supper currently separates its largest modes of customer-facing work, but those boundaries may evolve as the product develops.
Some functionality could combine, and new specialized agents could appear. Supper could eventually use a main-agent and sub-agent structure in which customers remain inside one broader conversation while specialized agents handle particular parts of the work.
The exact architecture is less important than the principle behind it: customers should not have to choose between the flexibility of conversation and the clarity of purpose-built software.
One interface does not require one agent. Sometimes the better conversational experience is an agent that knows exactly which job it is there to do.
Frequently Asked Questions About Supper's AI Agents
What is Supper?
Supper is an AI-powered business intelligence and analytics platform that helps people ask questions of company data in natural language while preserving the accuracy, governance, and visibility expected from a serious data stack. Supper combines conversational analytics, specialized AI agents, a business-aware semantic model, analytical tools, dashboard creation, and governed access to company data.
Why does Supper use multiple AI agents?
Supper uses multiple AI agents because conversational data analysis, semantic model maintenance, and dashboard creation are different workflows. Each requires different context, tools, goals, and definitions of a successful result.
What is Supper's Data Chat Agent?
Supper's Data Chat Agent, or DCA, is a general-purpose analytical agent. Customers use it to retrieve, manipulate, explore, analyze, and visualize company data through conversation.
What does Supper's Semantic Agent do?
Supper's Semantic Agent helps create and refine the semantic model that describes how a business defines and uses its data. It can investigate schema details, improve descriptions, and make business Terms precise enough to support reliable analysis.
What does Supper's Dashboard Builder Agent do?
Supper's Dashboard Builder Agent helps create datasets, charts, KPI tiles, and other visual components, then organizes them into coherent dashboards. It focuses on visualization choices, filters, themes, layouts, and relationships between dashboard components.
Does Supper replace data analysts?
Supper is designed to expand access to governed analytical knowledge, not remove the data team from the process. It helps employees answer more questions independently while allowing data teams to maintain the definitions, permissions, and analytical standards that make those answers reliable.
Is Supper a single chatbot?
No. Supper offers conversational interfaces, but the product is built around specialized agents and purpose-built analytical workflows rather than one chatbot responsible for every task.
One interface does not require one agent—and one chat box should not have to do every job.