Connect Orum to ChatGPT: Verify Accounts and Automate Transfers

If you need to connect Orum to ChatGPT to automate complex money movement workflows, verify enterprise bank accounts, or orchestrate subledger transfers, you need a Model Context Protocol (MCP) server. This infrastructure layer acts as the vital translation engine between ChatGPT's JSON tool calls and Orum's REST APIs. You can choose to build and host this integration middleware yourself, or use a managed integration platform like Truto to dynamically generate a secure, authenticated Orum MCP server URL.

If your team uses Claude, check out our guide on connecting Orum to Claude or explore our broader architectural overview on connecting Orum to AI Agents.

Giving a Large Language Model (LLM) read and write access to a sophisticated financial API like Orum is a serious engineering undertaking. You have to handle rigid scheduling logic for Orum Deliver, map asynchronous state machines for account verification, and navigate highly specific reference IDs for subledgers and external accounts. Every time Orum updates its schema or you want to expand the AI's capabilities, custom server code requires updates, redeployments, and exhaustive testing.

This guide breaks down exactly how to use Truto to generate a secure, managed Orum MCP server, connect it natively to ChatGPT, and execute complex financial operations using natural language.

Stop writing boilerplate API integration code. Let Truto generate secure, managed MCP servers for your financial AI agents in seconds. :::

The Engineering Reality of the Orum API

A custom Orum MCP server is essentially a self-hosted API gateway for your AI. While the open MCP standard provides a standardized way for ChatGPT to discover Orum operations, executing those operations flawlessly against Orum's highly regulated API surface is exceptionally painful.

If you decide to build an Orum ChatGPT integration in-house, you own the entire lifecycle. Here are the specific integration challenges that make building an Orum MCP server difficult:

The Subledger and Reference ID Web

Unlike standard SaaS CRUD APIs, Orum operates on deeply interconnected financial concepts. To initiate a transfer, you do not just pass a simple source and destination. You must manage customer_reference_id, account_reference_id, transfer_reference_id, and subledger_reference_id. If your LLM attempts to hallucinate a transfer payload, it will fail unless the MCP server strictly enforces JSON schemas that dictate these explicit Orum reference dependencies. Your tools must perfectly describe how a deliver_person relates to an external_account and how a subledger acts as a distinct pool of funds.

Orum's Strict State Machine Constraints

Orum resources adhere to immutable state transitions. For example, if you call delete_a_orum_deliver_person_by_id, the entity transitions to a closed status. It remains in GET responses for auditing, but cannot be reopened, and its reference ID can never be reused. Similarly, Orum Verify accounts rely on asynchronous processing (often involving micro-deposits). If you hand-code MCP tools, you must write extensive validation logic to prevent the LLM from attempting invalid state transitions on closed or pending Orum resources.

Complex Recurring Schedule Schemas

Orum Deliver supports recurring transfers via the schedules endpoint, which relies on strict RRULE strings (e.g., FREQ=WEEKLY;INTERVAL=1;BYDAY=MO). An LLM is prone to hallucinating invalid RRULE formats if not given highly constrained schemas. Your custom server must intercept the LLM's natural language intent, validate it against Orum's RRULE parser endpoint, and reject malformed logic before executing the schedule creation.

Factual Note on Rate Limits

When building an Orum ChatGPT integration, you must architect for scale. It is important to note that Orum applies its own API rate limits. Truto does not retry, throttle, or apply backoff on rate limit errors. When Orum returns an HTTP 429, Truto passes that error directly to the caller. However, Truto normalizes the upstream Orum rate limit information into standardized headers (ratelimit-limit, ratelimit-remaining, ratelimit-reset) per the IETF specification. Your client implementation is entirely responsible for interpreting these headers and executing its own retry or backoff logic. Do not assume the integration layer will silently absorb Orum rate limit rejections.

Orum to ChatGPT Quickstart Guide

If you just want the fastest path from a fresh Truto account to ChatGPT calling the Orum API, follow these steps. Deeper architectural and security details live in the sections below.

What you need:

• A Truto account with API access.
• Orum API credentials (API key or OAuth depending on your environment setup).
• A ChatGPT Pro, Plus, Business, Enterprise, or Education seat with Developer mode enabled.

Step 1: Connect Orum as an Integrated Account

In the Truto dashboard, navigate to Integrated Accounts -> New Integrated Account, select Orum, and provide the necessary API credentials. Truto securely manages the credential lifecycle, so ChatGPT never encounters an expired token.

Once connected, note your integrated_account_id.

Step 2: Generate the Orum MCP Server

Truto derives MCP tools dynamically from Orum's API documentation and your environment configuration. You can generate an MCP server URL either through the UI or via the API.

Method A: Via the Truto UI

1. Navigate to the Integrated Account page for your Orum connection.
2. Click the MCP Servers tab.
3. Click Create MCP Server.
4. Select your desired configuration (e.g., name, method filters, tags).
5. Copy the generated MCP server URL.

Method B: Via the Truto API Make a POST request to scope an MCP endpoint to your Orum account. You can filter by methods to restrict the LLM's capabilities:

curl -X POST https://api.truto.one/integrated-account/$INTEGRATED_ACCOUNT_ID/mcp \
  -H "Authorization: Bearer $TRUTO_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "name": "Orum for ChatGPT",
    "config": {
      "methods": ["read", "write"],
      "tags": ["deliver_transfers", "deliver_persons", "verify_accounts"]
    }
  }'

The API response returns a url field (e.g., https://api.truto.one/mcp/<token>). This URL contains a secure cryptographic token that handles routing and authentication. Treat this URL like a secret.

Step 3: Connect the Orum MCP Server to ChatGPT

Now, you must register the generated URL with your ChatGPT client.

Method A: Via the ChatGPT UI

1. In ChatGPT, navigate to Settings -> Apps -> Advanced settings.
2. Ensure Developer mode is enabled.
3. Under MCP servers / Custom connectors, click Add.
4. Name the connector (e.g., "Orum API Sandbox").
5. Paste the Truto MCP URL into the Server URL field and click Save.

Method B: Via Manual Config File (for local agents or Claude Desktop) If you are using a local agent or Claude Desktop, edit your JSON configuration file to use the Server-Sent Events (SSE) transport:

{
  "mcpServers": {
    "orum-api": {
      "command": "npx",
      "args": [
        "-y",
        "@modelcontextprotocol/server-sse",
        "https://api.truto.one/mcp/<token>"
      ]
    }
  }
}

Once connected, ChatGPT will negotiate the protocol handshake and list all available Orum tools.

Hero Tools for Orum Orum

By leveraging the documentation-driven approach, Truto surfaces Orum's endpoints as highly contextualized JSON-RPC tools. Here are 6 high-leverage hero tools your AI agents can execute out of the box.

Create an Orum Verify Account

Tool name: create_a_orum_verify_account

Description: Request account verification in Orum by submitting bank account details. This initiates Orum's verification engine to determine ownership, control, and debit status.

Usage note: The LLM must construct the payload using the account_number, routing_number, and account_holder_name. The response includes an estimated_verification_date and the initial verification_status.

"I need to verify a new external bank account for our vendor, Apex Corp. Their routing number is 121000358 and the account number is 555888333. Please initiate the verification process in Orum."

Create an Orum Deliver Transfer

Tool name: create_a_orum_deliver_transfer

Description: Initiate a money transfer in Orum between a source and destination party.

Usage note: The prompt must include a transfer_reference_id, amount, currency, and speed (e.g., RTP, FEDNOW, SAME_DAY_ACH). The LLM handles mapping the specific source and destination nodes required by the schema.

"Initiate a same-day ACH transfer for $4,500 from our main operational subledger to the verified external account for Apex Corp. Generate a unique transfer reference ID starting with 'APEX-INV-'."

List All Orum Deliver Businesses

Tool name: list_all_orum_deliver_businesses

Description: List all B2B customers in Orum, optionally filtering by customer_reference_id, status, or legal_name.

Usage note: Crucial for context retrieval. Before an agent can send money to a business, it uses this tool to find the business's internal ID and verify their status is not set to closed.

"Look up the business details for 'Apex Corp' in Orum. I need to confirm their account status is active and retrieve their Orum ID so we can process a payout."

List All Orum Balance Subledgers

Tool name: list_all_orum_balance_subledgers

Description: Retrieve all subledger account balances in Orum.

Usage note: Agents should call this before executing bulk payouts to ensure sufficient funds exist in the specific subledger, avoiding failed transfers.

"Check the current available balance on our 'Contractor Payouts' subledger in Orum to ensure we have enough funds to cover a $12,000 transfer run."

Create an Orum Deliver Schedule

Tool name: create_a_orum_deliver_schedule

Description: Create a schedule for recurring transfers in Orum Deliver.

Usage note: Requires strict RRULE compliance. The agent will formulate the RRULE, set the start_date, and map the source/destination accounts to automate recurring payouts.

"Set up a recurring weekly Orum schedule to pay $500 to John Doe's external account every Friday, starting this upcoming Friday. Use next-day ACH for the transfer speed."

Get a Reconciliation Report

Tool name: get_single_orum_reports_recon_by_id

Description: Get the status and metadata of a single reconciliation report in Orum by its ID.

Usage note: After an agent creates a reconciliation report request, it polls this endpoint to determine if the report generation is complete before attempting to download the CSV data.

"Check the status of the Orum reconciliation report ID 'recon-778899'. If it is completed, let me know the start and end times covered in the report."

To view the complete schema definitions and the entire tool inventory for Orum, visit the Orum integration page.

Workflows in Action

Connecting Orum to ChatGPT enables powerful, multi-step financial automations that eliminate manual treasury tasks. Here are two concrete examples of how an AI agent uses Orum MCP tools.

Scenario 1: B2B Onboarding and Account Verification

A treasury operations manager wants to onboard a new vendor and ensure their banking details are verified before issuing payment.

"Onboard a new vendor named 'TechSupply LLC'. Assign them a customer reference ID of 'TS-992'. Then, add their external bank account (Routing: 021000021, Account: 998877665) and immediately request Orum to verify it."

How the agent executes this:

1. Calls create_a_orum_deliver_business to register "TechSupply LLC" as a legal entity in Orum, passing the requested reference ID.
2. Calls create_a_orum_external_account using the ID returned from the first step to link the bank details to the newly created business profile.
3. Calls create_a_orum_verify_account to trigger Orum's account verification process on the newly attached external account.

Result: The user receives confirmation that the business and bank account were created, along with the estimated date that the verification process will conclude.

sequenceDiagram
    participant User as User Prompt
    participant ChatGPT as ChatGPT (MCP Client)
    participant Truto as Truto MCP Server
    participant Orum as "Upstream API (Orum)"

    User->>ChatGPT: "Onboard TechSupply LLC..."
    ChatGPT->>Truto: Call create_a_orum_deliver_business
    Truto->>Orum: POST /deliver/businesses
    Orum-->>Truto: Return Business ID
    Truto-->>ChatGPT: Result (Business ID)
    ChatGPT->>Truto: Call create_a_orum_external_account
    Truto->>Orum: POST /external_accounts
    Orum-->>Truto: Return Account ID
    Truto-->>ChatGPT: Result (Account ID)
    ChatGPT->>Truto: Call create_a_orum_verify_account
    Truto->>Orum: POST /verify/accounts
    Orum-->>Truto: Return Verification Status
    Truto-->>ChatGPT: Result (Verification Status)
    ChatGPT-->>User: "TechSupply LLC is onboarded and verification is pending."

Scenario 2: Automated Recurring Payout Validation

A finance admin wants to ensure sufficient funds exist before setting up a recurring weekly contractor payout.

"Check our main operating subledger balance. If we have more than $10,000, set up a recurring Orum schedule to pay $1,500 every Monday to Jane Smith's external account via RTP."

How the agent executes this:

1. Calls list_all_orum_balance_subledgers to query the real-time funds available in the operating subledger.
2. Analyzes the returned JSON. If the balance exceeds $10,000, it proceeds.
3. Calls list_all_orum_deliver_persons (or external accounts) to resolve Jane Smith's internal Orum IDs.
4. Calls create_a_orum_deliver_schedule using an RRULE of FREQ=WEEKLY;BYDAY=MO, mapping the source subledger to Jane's destination external account, setting the speed to RTP.

Result: The user is informed of the current subledger balance and receives the Orum schedule ID confirming the recurring transfer is active.

Security and Access Control

Giving an AI agent access to an enterprise Orum environment requires strict security perimeters. Truto MCP servers provide multiple layers of access control configured at server generation:

• Method Filtering: Limit the Orum MCP server to specific operations. By passing methods: ["read"], you can allow an agent to query subledger balances and view reports without granting it the ability to execute Orum Deliver transfers.
• Tag Filtering: Restrict tools by domain. If you only want an agent managing verifications, configure tags: ["verify_accounts"] to hide all Orum Deliver and Webhook tools.
• Require API Token Auth (require_api_token_auth): When enabled, possessing the Truto MCP URL is not enough. The client (e.g., ChatGPT) must also inject a valid Truto API token in the Authorization header, enforcing secondary authentication.
• Auto-Expiration (expires_at): You can generate ephemeral Orum MCP servers for temporary audits or short-lived agent tasks. Truto automatically cleans up the server and its keys when the timestamp is reached.

Scaling Agentic Finance with Orum and Truto

Building an Orum ChatGPT integration from scratch forces you to manage dynamic JSON schemas, interpret rigid financial references, and handle deep state machines. Every time Orum adds a new feature or rate limit behavior, your custom middleware breaks.

By leveraging Truto's dynamic MCP server generation, you offload the entire infrastructure burden. Truto instantly converts Orum's API documentation into LLM-ready tools, maintains secure token lifecycles, and standardizes authentication. This allows your Orum ChatGPT integration to go live in minutes instead of months, empowering your AI agents to automate treasury operations, verify enterprise accounts, and manage subledgers safely and accurately.