Infinite Audience MCP

Connect your own AI models and agents directly to the graph via the Model Context Protocol. Native tools for identity resolution & enrichment, building segments, and activating audiences.

Model Context Protocol Server Architecture

Universal MCP Compatibility

Standard Model Context Protocol features native to any compliant AI client.

Resources

Context & Schemas

Stream the complete Infinite Audience Knowledge Catalog straight to your model. Deep graph metadata provides the exact structural context needed for flawless, hallucination-free query construction.

Tools

Active Operations

Mount active operations directly inside your model's execution environment. Connect agents to real-time identity resolution and enrichment pipelines, or execute analytical and segment-creation queries on demand.

Prompts

AI Templates

Embed production-ready prompt templates and task configurations straight into your AI environment. Pre-configured message structures and native variables provide the baseline constraints needed to safely trigger autonomous, agentic workflows in minutes.

Cognitive Operations. Seamless Integration.

Bridging the gap between raw data and cognitive reasoning loops to provide agentic intelligence.

Dynamic A2A (Agent-to-Agent) Workflows

Unlike a static UI dashboard that requires human clicks, or a traditional API that requires rigid code, MCP gives AI agents native "hands" to query the graph. Your autonomous workflows can dynamically seek out audience data, resolve identities, and activate campaigns in response to real-time goals.

Tool-Calling Integration

Conversational Insights & Reasoning

AI models connected via MCP can run complex cohort analyses on the fly. You can ask an agent to "analyze last week's cohort, identify the top three demographic attributes, and outline the overlap with our primary segment." The model executes the tools, parses the graph, and explains the results in plain language.

Natural Language to SQL (NL2SQL)

Self-Optimizing Campaigns

Because AI models can evaluate their own outputs, an MCP-connected agent can monitor matching rates during data enrichment. If match rates are lower than expected, the agent can dynamically adjust columns, clean formatting, or refine filtering rules, self-correcting the pipeline in real-time.

Feedback loop orchestration

The B2C Agent Execution Lifecycle

See how an autonomous agent uses MCP tools behind the scenes to orchestrate a complete identity resolution and activation workflow.

Tool: audience_lookup

Secure Lead Match

The client agent queries audience_lookup using identifiers (e.g. hashed email or phone number) to search the graph. This matches the user to a canonical consumer record inside the secure graph.

Tool: enrich_match

Context Enrichment

Using the matched record identifiers, the agent invokes enrich_match to JOIN 525+ premium demographic, lifestyle, and buying intent attributes, loading complete consumer context into the model's environment.

Tool: audience_overlap

Conversational Analytics

The agent calculates audience size, overlap indices, and attribute breakdowns by calling audience_overlap. This allows LLMs to run complex target market analysis via conversational queries.

Tool: purchase_audience

Direct Platform Push

Once segments are dynamically compiled or curated by the model, the agent executes purchase_audience to trigger the secure onboarding sync to downstream DSP and SSP ad seat configurations.

Beyond the Dashboard: Choosing Your Surface

Our platform provides three distinct ways to interact with the Infinite Audience Graph. Here is how they complement each other to fit your team's workflow:

Surface 01

Visual UI Platform

A point-and-click dashboard designed for marketing managers and analysts. Perfect for building visual segments, uploading CSV files manually, and viewing analytics.

Best For: Human-driven campaign planning & manual uploads.

Surface 02

Programmatic API

Low-latency REST endpoints for developers and systems. Enables real-time profile lookup at checkouts, automated customer syncs (CRM/warehouse), and system-to-system integrations.

Best For: Backend software engineers & CRM data automation.

Surface 03 (This Page)

Model Context Protocol (MCP)

Gives AI models and cognitive agents native "hands" to query and edit the graph. Supports natural language queries, autonomous optimization, and dynamic agent workflows.

Best For: Autonomous AI agents & conversational LLM reasoning.

Expose Infinite Audience to AI Agents

Request access to our Developer Portal to obtain sandbox keys, explore the MCP server spec, and download client configurations.

Request Developer Access