Case Study

Brian Impact Foundation powers their search for the next generation of social innovators with Claude

Brian Impact logo

Brian Impact Foundation, a nonprofit focused on using technology to solve social problems, used Claude to help the team review 100x more candidates than the year prior, allowing the team to consider a much larger pool of innovators and create a more equitable process for identifying fellowship candidates.

Large impact, small team

In 2021, Kakao founder Brian Beom-su Kim pledged to give more than half his $8.9 billion net worth to social causes. He launched Brian Impact Foundation to support individuals and organizations using technology — especially AI — to solve social problems. Today, Brian Impact Foundation awards grants to nonprofit organizations and supports individual innovators through fellowships.

But like most nonprofits, the foundation operates with a lean team. Just 13 people manage millions of dollars in grants. “Even though you try your best to go through every candidate in a very equitable way, you run out of time and resources,” says Soomin Syd Kim, project director at Brian Impact Foundation.

Generative AI as a lever for scale

Soomin’s team turned to generative AI to help the organization expand its fellowship applicant pool and assess each candidate more thoroughly. They built a platform they call Benevolence Enhancing Expert System, or BEES, which uses Claude to turn massive amounts of data into an easy-to-read summary about potential fellows.

“Most of the data we use in the philanthropy sector is unstructured, so it is perfect for GenAI and large language models,” Soomin says, noting that it’s a similar problem faced by investors performing due diligence.

When selecting an LLM, she was drawn to Anthropic’s focus on safety, as well as Claude’s capabilities. “Claude’s long context window was significant for us because we process a lot of news articles,” she says.

Accelerating application review with Claude

The team uses BEES for two use cases: identifying prospective fellowship candidates and reviewing fellowship applications.

BEES first gathers diverse types of data related to social problems and innovation from the open web. It sends the raw data to Claude to extract potential candidate information, summarize the contents, and analyze the quality, structure, and type of the data. Then, using information that meets a certain quality threshold, Claude organizes the information based on the criteria the foundation uses to assess applicants.

Information is surfaced to reviewers in a custom web app designed to make it easy for them to review everything about the prospective candidate or applicant: various kinds of information that might help Brian Impact Foundation understand their work as an innovator.

Previously, reviewers had to conduct laborious manual searches to gather information about each prospective fellow. “This is really huge in terms of improving efficiency in resource allocation,” Soomin says.

How Claude drove a 100x increase in candidates

Prior to selecting Anthropic for foundation models, Brian Impact Foundation relied on members of its expert network to recommend candidates for fellowships. The team knew they were potentially missing out on supporting innovators who could benefit from the foundation’s support, but they simply didn’t have the time or resources to broaden their search.

Last year, with Claude, they reviewed over 20,000 candidates—one hundred times more candidates than the previous year.

Claude also accelerates the process of reviewing fellowship applications, enabling the team to study every applicant deeply and select fellows based on the most thorough review possible. “We tripled the number of reviews for each candidate,” Soomin says.

With Claude, Brian Impact Foundation can be confident that it’s supporting the most deserving social innovators without diverting resources away from its core mission: creating positive change.

“Partnering with Anthropic has helped our team of 13 discover a larger number of fellowship applications and review every application more thoroughly. This enables us to increase equity in our fellowship application process, strengthening our mission to support innovators using technology to make the world a better place.”

- Soomin Syd Kim, Project Director at Brian Impact Foundation