%201.webp)
Impala has launched an MCP server that brings its philanthropic data into Claude and ChatGPT, aimed at foundations, grantmakers and nonprofits.
The product gives users access to data on funders, nonprofits, grant histories, governance, financial health and funding flows inside AI tools. Organisations can query the information in natural language or use it to build their own applications and workflows.
The server is based on the Model Context Protocol, or MCP, an emerging standard for connecting AI models with external data sources and software tools. Impala said its offering is among the first in the philanthropy sector to use that architecture and that it is listed in Anthropic's Connectors Directory.
The launch reflects a wider shift in how AI is being used across philanthropy. Much of the recent focus has been on tools that help users draft grant proposals, but Impala is targeting the data layer behind grantmaking decisions, due diligence, portfolio analysis and funding strategy.
Its platform provides structured information on individual organisations and broader funding ecosystems across geographies and causes. This includes details on leadership, governance structures, grant histories, financials, personnel, trends over time and funding gaps.
Users can also combine Impala's dataset with their own systems, including grant management software, fundraising customer relationship management tools and internal data. That allows organisations to build more tailored dashboards, analysis tools and decision-support systems inside Claude Code and OpenAI Codex, as well as general chat interfaces.
Sector use
One early user is the Jewish Federation Bay Area, where teams had spent years moving across multiple platforms to answer questions about their grant portfolio and related organisations.
"The questions that foundations and nonprofits need to answer are effectively endless, and they're different for every organisation," said Marisa Sires, Chief Product Officer at Impala.
"Our MCP makes it possible to leverage AI and tailor applications to how an organisation operates. That means a board reporting tool built around your grantees, a due diligence agent designed for your investment thesis or a portfolio analysis workflow calibrated to your specific strategy. Now, the full depth of sector-wide philanthropic intelligence is available to power whatever an organisation needs to build," Sires said.
Justin Stevens, Managing Director of Data and Business Operations at Jewish Federation Bay Area, said the new approach had changed how his team worked with sector information.
"Impala's MCP changes how our team accesses and applies philanthropic intelligence," Stevens said.
"With this level of data at our fingertips, we're no longer toggling between platforms. Instead, we're able to answer specific questions that are tailored to our needs with meaningful intelligence and data. For a foundation managing a complex portfolio, that kind of flexibility is genuinely transformative," he said.
Data layer
Impala's product pitch centres on the idea that philanthropic organisations often need bespoke analysis rather than standard software outputs. A foundation may want to compare its grantmaking with peer funders, assess the financial health and governance of a prospective grantee, or map funding gaps in a particular region or issue area.
In that context, the company is positioning its serverless as a standalone application and more as infrastructure that supports a range of AI tools. The aim is to let users ask specific questions against a verified dataset rather than rely only on a large language model's general knowledge.
Impala said its data includes IRS-verified philanthropic information. Customers can identify funders aligned to particular causes and geographies, benchmark their strategies against other grantmakers and analyse networks of organisations and donors.
For some organisations, Impala is also offering service engagements to build custom AI agents and applications. That creates a second line of business beyond direct access to the MCP server itself, combining software access with implementation work for clients that do not want to build their own tools.
The server is available immediately for Claude and ChatGPT users. Anthropic added the product to its directory after a review for security, reliability and compatibility with Claude products.
Impala said its dataset is already used by organisations including the Robin Hood Foundation, Carnegie Corporation of New York and the Minnesota Council on Foundations.