Philanthropy data belongs where you already work

Introducing the Impala MCP: the AI data layer for philanthropy. Here is what it is, why we built it, and how to start using it in about two minutes. 

For years, the people doing the most important work in the world have been handed the least impressive tools.

Think about who we are talking about. A program officer weighing where a foundation's next million dollars should go. A development director trying to understand a funder before the first meeting. A researcher mapping who else is working on the same problem in the same city. These are high-stakes, high-judgment jobs. And for a long time, the software built for them lagged years behind what every other sector took for granted.

We started Impala to close that gap. We built the most complete, most trustworthy data platform in philanthropy, covering more than two million nonprofits, over 250,000 foundations, 16 million board members and leaders, and tens of millions of grant transactions. But a platform was never the whole point. The point was always to meet the people in our field where they actually work, with tools that fit the way they think.

Today, more and more of that work happens inside AI assistants. So that is where we are putting the data.

Introducing the Impala MCP

The Impala MCP connects Claude, ChatGPT, Gemini, Copilot and any other AI assistant directly to Impala's verified philanthropy data. You ask a question in plain English, and your assistant answers from real data on nonprofits, foundations, and grants. No new tab. No exporting a spreadsheet. No toggling between platforms while you lose your train of thought.

MCP stands for Model Context Protocol, an open standard that lets AI assistants securely reach trusted sources of data. For you, it means the assistant you already use becomes fluent in philanthropy, grounded in data you can verify.

Ask "Which foundations in Ohio fund youth mental health programs, and how much did they give last year?" and you get an answer built from real filings, with the source attached. Ask your assistant to profile a funder before a meeting, draft a board-ready memo on a grantee, or map every organization working on food security in your region, and it does the work with you, in the same window where you already write, plan, and think.

What you can build on day one

You're not limited to the reports someone else decided to build. Because the data lives inside your assistant, you can shape it around your own questions and augment it from data coming from other systems (i.e. GMS, CRM). A few examples of what teams can build:

• A development team builds a funder shortlist, ranked by fit and recent giving, in the time it used to take to open the research.
• A program officer turns a grantee's filings into a board-ready due diligence memo, complete with financial health signals and a clear summary of risk.
• A strategy lead produces a landscape map of every funder and nonprofit working on an issue, so they can see the whole field before deciding where to move.
• A portfolio manager pulls risk signals across a set of grantees, flagging the organizations that need a closer look.

You can build these yourself, or you can tell us what you wish existed, and we will build it with you.

Included with your plan. No new bill.

The MCP is included with your Impala subscription and follows whatever plan you're already on.

On a free Impala account, you get access to Impala Essentials: discovery across more than two million nonprofits and 250,000 foundations, financials, 16 million board members and leaders, core profiles and cause areas, 990 summaries, and a fair-use limit generous enough for real research. It costs nothing.

If you subscribe to Impala's Philanthropy Database, your assistant reaches everything in the premium tier too, including more than 25 million individual grant transactions, full multi-year 990 line items, advanced giving analysis, and funder overlap. Same data you see on the platform, now inside your assistant, with no add-on fee.

Built to be trusted

When the answers inform where money goes and who gets funded, trust is not a feature. It is the foundation. So we designed the MCP to earn it at every layer.

It is read-only by design. Impala can retrieve, never write. It cannot create, edit, or delete anything in your account, your lists, or your CRM.

Access is authenticated and permission-aware. There is no anonymous access. You sign in with your Impala account, and your entitlements are enforced on our servers, so an assistant can never reach data your seat cannot.

Your prompts are never used for training. Your queries serve your request and a private audit log for your account. They are never used to train models, ours or anyone else's.

And every answer is verifiable. Results trace back to IRS filings and organization-verified profiles, and each one carries an EIN and a link back to its Impala profile, so you can check the source before you act on it.

The first step in a bigger plan

The MCP is the first move in building Impala into the AI platform for philanthropy:

A place where the people who fund and run social-impact work can bring any question they have and get a trustworthy, actionable answer, in whatever tool they already use. This launch is the beginning of that, not the end.

The whole sector, behind every question you ask.

Try it yourself

Setting up takes about two minutes. Create a free Impala account, add the connector in Claude or ChatGPT, and start asking.

Get started for free at www.impala.digital/mcp

If you're interested, we're also hosting a live webinar to explain how to use the MCP: from initial setup to complete workflows and integrations. It's happening on September 16 at 1:00 ET (at no cost), and you can register using this link.