Connect any MCP-compatible AI assistant to the AIARCO Ads Engine. Collect behavioral signals from conversations and deliver contextual offers — all through a single hosted server.

Why Use MCP?

The Model Context Protocol is an open standard that lets AI assistants connect to external tools. Instead of building custom REST integrations, your AI client automatically discovers AIARCO tools and invokes them during conversations.

Without AIARCO MCP

  • Manual REST API integration - Custom signal extraction logic - Build your own offer rendering - Handle tracking separately

With AIARCO MCP

  • Zero-config tool discovery - AI extracts signals automatically - Contextual offers in-conversation - Built-in impression tracking

How It Works

  1. Connect — Point your MCP client to https://mcp.aiarco.com/mcp
  2. Signal — The AI assistant extracts behavioral signals (interests, purchase intent, brand affinities) from conversations and sends them via broadcast_signal
  3. Offer — Call get_offers to retrieve contextual, personalized ad offers matched to those signals
  4. Track — Report viewability with track_impressions to complete the loop

Key Features

  • Passive Signal Collection — Signals are extracted from natural conversation, not explicit user actions
  • 9 Signal Categories — Interest, evaluation, purchase intent, brand affinity, price sensitivity, and more
  • Contextual Matching — Offers are matched in real-time based on signal context, not keywords
  • Privacy-first — No PII required; optional hashed identifiers for frequency capping
  • MCP Standard — Works with Claude, GPT, VS Code Copilot, Cursor, Windsurf, and any MCP-compatible client

Available Tools

ToolDescription
broadcast_signalSend behavioral signals extracted from conversations to the Ads Engine
get_offersRetrieve contextual ad offers based on a query and optional signals
track_impressionsReport impression viewability after offers are displayed to the user

Next Steps

Setup & Configuration

Connect to the hosted MCP server in under 2 minutes.

Tool Reference

Schemas, parameters, and examples for each tool.