Euterria

Taken together, the Bay Area climate community, as is the case with most regional networks of nonprofits, knows an enormous amount. There are years of renewable energy pilots, program designs for home retrofitting, educational materials, and hard-won lessons about what works. But community-based organizations have little visibility into what other organizations around them are doing, even when that work is highly similar. The desire to know the lay of the land, to see who’s working on what, and perhaps to identify overlap that could spark collaboration is certainly present. Most people running these organizations are principally motivated by the desire to do right by their community and the planet, and they easily see why collaboration can help meet that end.

But the collaboration cost is too high. Digging up the old report, traveling to meet someone, and working out who owns which part of a service all take time and money these organizations don’t have. Moreover, these organizations are often competing for the same pool of money. So while the collaborative spirit tends to guide personal interactions between members of various organizations, the result of this funding environment is a lack of incentive to collaborate.

My reference point for this is academia, which has plenty of its own problems, but does one thing very well: leaving a paper trail. Publishing makes the state of a field legible, and funding, explicitly or implicitly, is tied to how a project builds on and advances what’s already been done. A new researcher can spend a week or two reading papers and be relatively up to speed on which problems the field is most focused on. In the private sector, a market does similar work: a duplicate product loses to the incumbent, and the incumbent loses to whoever ships something better.

Neither of those correcting forces exists for nonprofits. There’s no market forcing convergence and no funding structure that specifically rewards knowing what already exists and building on it.

In sum, there’s a knowledge problem, where organizations struggle to know what other organizations are working on, and an ecosystem problem, which is the lack of collaborative incentive.

Euterria, which I co-founded in 2025 with early backing from Stanford’s TomKat Center for Sustainable Energy, is the platform we are building to fix the knowledge problem and serve as the foundation for a more collaborative Bay Area climate ecosystem.

Making the field legible to itself

Euterria is a nonprofit knowledge-sharing platform for the Bay Area climate community: a shared resource built for sharing resources. Organizations drop in their URL and upload what they already have, decks, reports, PDFs, and Euterria turns those documents into an internal search engine for the whole network, making the community’s programs, people, and resources easy to discover and actually use. Then anyone can query it the way they’d ask a well-connected colleague, “who’s working on wildfire smoke and indoor air quality?”, and get a synthesized answer with the actual organizations, initiatives, and people to reach out to.

A starting point

Euterria has already proved useful to 30+ organizations around the Bay Area, and is growing every day as organizations add more material. But we’re well aware of the challenges. Keeping data fresh and users coming back to a platform designed to be useful are Euterria’s central obstacles. To really solve the collaboration problem for nonprofits, we have to focus on the primary needle-mover for nonprofits: money.

Nonprofits are service providers, and it’s often necessary to have a mesh of organizations serving similar needs across different populations and regions. I don’t advocate for a market of specialized nonprofits as a means of creating efficiency as I would in a competitive market. But I think funders have an interest in understanding where the nonprofit space is in a given region, and who has developed real-world know-how, in order to best allocate funds to uplift the region as a whole, not just a particular organization’s work. Euterria’s platform is a starting point for building the capability of knowing where the field stands, enabling funders and organizations to have a clear picture of who’s working on what, and creating a substrate for low-friction collaboration among organizations that can be incentivized to work together through grant structure.

Technology

You can read more about the search architecture here if you’re interested in the specifics. But broadly speaking, slide decks and PDF reports are human artifacts and thus messy, context-heavy, and rarely shaped like database records. So the first thing Euterria does with an uploaded document is read it with a model and pull it apart into typed records: a program, with its issue areas and the people it serves; an organization, with what it needs and what it offers; a person, with their skills and role. Long documents get split into sections and indexed on their own. Nothing stays a free-floating blob of text. Everything becomes an entity with fields you can point at.

When someone asks Euterria a question, we use a reranker under the hood. Each candidate, whether a person, a program, an organization, or a passage from a report, gets rendered into a compact semantic document: its title, its type, the organization behind it, its metadata, and a snippet or two. Our reranking model reads all of them against the original question and reorders the shortlist by what each result actually means, not by how often it echoed the query’s words. This is critical to surfacing who is actually the most useful to talk to, since organizations frequently use different vocabulary to describe similar problems.

Keyword shortlist

1. 1Heat & Health Report
2. 2Citywide Tree Census
3. 3Oakland Shade Plan
4. 4Dr. Maya Okonkwo
5. 5GreenLine Collaborative

semantic reranker

Reranked by meaning

1. 1Dr. Maya Okonkwo▲ 3
2. 2Oakland Shade Plan▲ 1
3. 3GreenLine Collaborative▲ 2
4. 4Heat & Health Report▼ 3
5. 5Citywide Tree Census▼ 3

Person Program Organization Resource

Piloting with 20 organizations across the Bay

As of June 2026, there are 20 organizations across the Bay uploading their own material and sharing with us what search gets wrong. We pay these organizations for their time and materials because their feedback is so valuable as we build the best platform possible.

A platform that grows in conjunction with the collaboration culture

A search box can surface who’s working on what, but it can’t make two organizations pick up the phone or manufacture the trust that’s earned by being an excellent nonprofit operator over the years. I do believe the right technology can lower collaboration friction. This is the essence of why businesses spend millions per year on software to help their employees work together. But the tech ought to be accompanied by a regional culture that values collaboration and knowledge sharing as a means of better serving the community.