OpenAI / ChatGPT + Connect AI
Build GPT-powered data assistants using the OpenAI Python SDK with live data from hundreds of sources via Connect AI’s MCP server.
from openai import OpenAI
client = OpenAI()
# MCP tools expose Connect AI data sources
# as OpenAI function definitions
tools = mcp_client.list_tools_as_openai()
response = client.chat.completions.create(
model="gpt-4o",
tools=tools,
messages=[{
"role": "user",
"content": "Which accounts have open opps over $50k?"
}]
)
# Handle tool calls, execute against live data
Developer guide
Full step-by-step guide to building a production-ready GPT data assistant with Connect AI MCP.
Sources for OpenAI
Common GPT + source patterns. Tap any combination to see the tutorial.
Sample agents
Two starting points: a one-command scaffold, or a deeper reference implementation.
Generate a working LangChain, Google ADK, or Vercel AI SDK agent (OpenAI-backed) in one command. Includes Management MCP setup for in-band connection provisioning.
Production-ready OpenAI data assistant with Connect AI MCP, function calling, and multi-turn conversation context.
Evaluating Connect AI for your team? See the full OpenAI integration overview.
Full API reference, authentication guides, and configuration details.