Can ChatGPT Access Institutional Investor Data?

Can ChatGPT Access Institutional Investor Data?
5:05

Data sourced from Dakota Marketplace, the global LP and GP intelligence platform trusted by thousands of investment professionals. Learn More | Book a Demo

Ask ChatGPT for a family office's AUM, its investment preferences, or a portfolio manager's direct contact, and it will guess. Generic AI models are trained on public web text, not verified institutional investor data, so the model fills the gap with something plausible-sounding and often wrong. For a fundraising team about to send an email off that name, wrong is expensive.

OpenAI has now approved Dakota Marketplace for the ChatGPT app store, and Dakota's verified allocator, account, and contact data can be queried directly inside a ChatGPT conversation.

What's Now Available

Fundraisers already use ChatGPT to draft emails, prep for meetings, and research firms. Until now, they had to leave that conversation and log into Dakota Marketplace separately to check the underlying data. Once a member connects their Dakota Marketplace account inside ChatGPT's Apps directory, that switch goes away. Dakota's data sits alongside everything else in the chat.

Inside a connected conversation, a user can ask for a list of family offices in a specific state with allocations to a given asset class, or a contact summary for a named portfolio manager, and Dakota's verified, daily-updated records answer the question instead of ChatGPT's general training data.

The connection covers Dakota Marketplace's full allocator universe: more than 173,900 accounts and 627,900 contacts globally as of January 2026, spanning RIAs, corporate pensions, family offices, insurance companies, public pensions, and international allocators across 91 countries.

One access note: this currently requires a ChatGPT Enterprise or ChatGPT Business subscription. The Apps directory isn't available on the free or Plus tiers.

How It Works With Dakota Marketplace

The ChatGPT app isn't a one-time export sitting inside a chat. It's a live connection back to your Dakota Marketplace account. When the app is attached to a conversation, every question you ask gets answered from Dakota's actual account, contact, and investment preference records, updated daily, rather than from ChatGPT's general training data.

That distinction matters because it's the difference between a guess and a sourced answer. Ask base ChatGPT about a family office's allocation history and it will produce something plausible-sounding whether or not it's accurate.

Ask the same question with Dakota Marketplace attached, and the answer is pulled from the same verified records you'd see logged into Marketplace directly, covering account intelligence, contacts, investment preferences, and, depending on the accounts involved, plan minutes, fee data, or Form D filings. It's the same grounded-data approach behind Dakota's existing Claude connection, now extended to ChatGPT.

What to Ask It

The app is built for the questions fundraisers ask most, phrased however you'd naturally type them:

Prospecting and list-building

  • "List family offices in Texas with allocations to private credit."
  • "Which public pensions in the Southeast have made a new manager hire in the last quarter?"
  • "Show me RIAs in the Northeast with over $500 million in AUM that invest in liquid alternatives."

Account and contact research

  • "Give me a contact summary for [portfolio manager name]."
  • "What are [allocator name]'s investment preferences, and who's the right person to contact there?"
  • "Has anyone at [firm name] changed roles in the last six months?"

Meeting prep and outreach

  • "Pull everything Dakota has on [allocator name] before my call with them tomorrow."
  • "Draft a first-touch email to [contact name] referencing their fund's recent allocation activity."
  • "Summarize this account's history with our firm, if there's a record of prior contact."

Each of these runs against Dakota's live data, so the answer reflects what's in Marketplace right now, not a cached snapshot from whenever the model was trained.

How to Connect It

  1. Open ChatGPT and go to the Apps directory.
  2. Find Dakota Marketplace and click Connect.
  3. Sign in with the same email and password used to log into Dakota Marketplace.
  4. Once connected, the button changes from Connect to Start chat.
  5. Click Start chat, attach Dakota Marketplace to the conversation, and start asking.

This requires a ChatGPT Enterprise or ChatGPT Business subscription. The Apps directory isn't available on the free or Plus tiers. For the full walkthrough with screenshots, see how to connect Dakota Marketplace to ChatGPT.

More than 173,900 accounts and 627,900 contacts, across every allocator type, U.S. and international, are now one connected ChatGPT conversation away. See what Dakota Marketplace looks like inside ChatGPT, inside Claude, or inside your CRM. Book a demo to get connected.

Cate Costin, Marketing Associate

Written By: Cate Costin, Marketing Associate