Captação com Causa: what we learned automating a fundraising program for CSOs

A fundraising program saves time when technology does the repetitive reading of calls for proposals and the team keeps the mentoring. In Captação com Causa, AI extracts the call's summary sheet, answers eligibility questions and suggests project ideas, and the program's people keep following each organization.

What Captação com Causa is

It is an initiative of Fundação FEAC (FEAC Foundation), supported by Instituto Phomenta and CORRE.SOCIAL, created so that NGOs in Campinas can diversify their income strategically. According to the program page, participants get training, individual and group mentoring, and access to AI tools to design projects, raise funds through calls for proposals and report on results.

How it was: each call read by each organization

A small NGO that wants to raise funds has to answer, before writing a single line, whether it can apply to the call. That means reading a long document, finding the eligibility item and the list of documents, and checking them against its own situation. When each organization in the program does this alone, the same call is read many times and interpreted in different ways.

How it is now: three blocks

Call summary sheet extracted by AI. The AI reads the call and fills in a sheet with funder, dates, amount, areas and target audience, plus the eligibility requirements and the required documentation. The organization sees the essentials before deciding whether it is worth reading the rest.

A Q&A assistant that knows how to say no. On the program screen, an association with two years of CNPJ (Brazil's tax registration) asks whether it can apply, and the assistant answers that it cannot, citing item 7.1 of the call. An assistant that only says yes makes the organization lose weeks writing a proposal that will be eliminated. Here, the answer is anchored in the text of the call.

Tailored project ideas. The AI suggests project directions tailored to the organization. It is a starting point for the conversation, and the final design stays with the organization's team and with the mentoring.

The voice of someone who used it

One participant, Letícia, from PROGEN, said about the writing robot: "It was relevant for improving the quality of our proposals and for understanding better how to structure a project" (our translation). What stands out is what she highlighted: understanding how to structure a project, and not only finishing faster.

What we learned

• Anchor the answer in the document. Citing the item of the call makes it possible to check and to trust.
• Teaching it to say no is part of the product. A negative answer at the right moment saves more time than any draft.
• AI writes drafts, people decide. The sheet and the ideas are a starting point for the organization and for the mentoring.
• The program is still people. Technology takes the repetitive reading, and the training and follow-up stay with people.

The full Captação com Causa case has the project design. The same design works for anyone who supports several organizations and needs to follow them at scale. See the page Networks and programs that listen and the Lab Nova Longevidade case for another program of this kind.