ChatGPT connectors for business data

Connectors you can use from ChatGPT

One Nexatron connection reaches every system your workspace connects, and every connector below is available. Where a system supports it, each person signs in with their own account, so ChatGPT sees only what they may see.

Questions teams bring to ChatGPT

• SAP Ariba spend analysis in your own words
• SAP Concur expense reporting in your own words
• Coupa spend analytics in your own words
• ServiceNow SLA reporting in your own words
• NetSuite month-end close questions in your own words
• Salesforce and NetSuite in one question
• Salesforce and SAP S/4HANA in one question
• HubSpot and QuickBooks in one question
• Snowflake and Salesforce in one question
• ServiceNow and Workday in one question
• SAP Ariba and SAP S/4HANA in one question

ChatGPT connectors FAQ

What are ChatGPT connectors?

Connectors let ChatGPT use tools and data from other applications, and custom connectors use the Model Context Protocol (MCP). Nexatron’s hosted MCP server works as a custom connector: one connection gives ChatGPT governed access to the systems your Nexatron workspace has connected.

Which ChatGPT plans can add Nexatron?

ChatGPT can add remote MCP servers as custom connectors on some plans, when the workspace allows it. On business plans, a ChatGPT workspace admin usually controls this. If you see no option to add a custom connector, ask your ChatGPT workspace admin, and check OpenAI’s current help pages for the exact menu labels.

What can ChatGPT use once Nexatron is connected?

The same hosted tools as any other client: 7 analytics tools, plus 3 action tools when the workspace allows them and the connection was granted an action scope. Every call runs as you, with your Nexatron role, row policies and masking applied.

Which business systems can ChatGPT reach through Nexatron?

Every connector listed on this page is available from ChatGPT. When each person signs in as themselves, answers respect the permissions set in that system. A shared connection reads the same data for everyone enabled on it, and access rules set in Nexatron decide who can use it.

Can one ChatGPT question combine several systems?

Yes, where the systems are connected. Nexatron queries each source within the asking person’s access and combines the results in an isolated step inside your workspace, so a question can set Snowflake usage beside Salesforce opportunities without a warehouse migration. The answer shows the generated query for each source and how many records matched, and these answers take longer than single-source ones.

Whose permissions does an assistant use?

The asking person’s. Each person signs in with their own Nexatron identity through OAuth 2.1 with PKCE, so the assistant sees only what that person may see in Nexatron and, for per-user connectors, in the source system itself.

Can an assistant change data in our systems?

Questions never change data. A separate set of action tools can, for connectors that support actions: nothing can be written until a workspace admin turns on write-back for that connection. Each action is previewed and needs explicit confirmation by a permitted person. An admin can also remove the action tools altogether.

Do we need a Nexatron workspace first?

Yes. The MCP server answers from a Nexatron workspace with at least one connected source, and a workspace administrator approves Nexatron for the workspace.