The AI Philanthropy Team
Every draft an agent produces waits in the Approval Queue for Alyssa. Nothing is posted or sent automatically. Click an agent to view and edit its specification.
Grant winnability
Athena's own 1 to 5 estimate of how likely Inheris is to win each foundation grant, learned from the applications she has already won. It is a measure of her effectiveness as a grant writer, tracked over time against actual outcomes. Winnability lives with Athena on the Donor System; this tab reads it. She estimates each grant there.
Loading winnability...
Scout Research: nonprofits and foundations (ProPublica)
Scout works the IRS 990 data on its own: it plans its own searches from the mission and the pipeline's gaps, pulls each organization's 990, scores donor fit, and queues the best here, tagged and ranked. Nothing enters the pipeline without your approval.
Scout Research: donors and committees (OpenFEC)
Scout works the FEC campaign-finance data on its own: it plans its own searches, pulls individual donors and committees, scores donor fit, and queues the best here, tagged and ranked. Public record, read as a capacity and generosity signal, not a political profile. Nothing enters the pipeline without your approval.
Scout: evaluate a lead
Paste a bio, post, or article about one person or organization. Scout judges mission fit, scores it, and can add a fit to the pipeline. Scout does not scrape any platform.
Scout: ProPublica
Look up a specific nonprofit or foundation in the IRS 990 dataset (free, public). A foundation that files a 990-PF is a grantmaker, and its assets are its giving capacity. Scout assesses fit on the real 990 financials and can add it to the pipeline.
Scout: OpenFEC
Look up a specific individual donor or committee in the FEC campaign-finance data (public record). Scout reads giving as a capacity and generosity signal, assesses fit on the real numbers, and can add it to the pipeline.
Scout: vendor feed
Pull leads from a licensed prospect-data vendor (iWave, DonorSearch) through Scout's adapter. Not yet connected: the mock vendor returns clearly fabricated placeholder leads until a real vendor is wired in behind the same adapter.
Buster: social candidates
Buster reads PUBLIC posts on mission topics from Bluesky and Mastodon (both open protocols) and surfaces people who look like potential contacts. Review each, then add the good ones to the pipeline or dismiss. Public data only, nothing is posted, and nothing reaches the pipeline until you add it.
Add a prospect
Pipeline by stage
Engagement order · click a stage to filterPipeline
| Name | Fit | Stage | Cares about | Brief | Donor (IDS) | Updated | |
|---|---|---|---|---|---|---|---|
New campaign
Campaigns
Ali: the day's moves
Ali reviews the pipeline and suggests one next step per priority prospect. She proposes; Alyssa approves.
Gift milestones
When a promoted prospect gives, the team reacts: a thank-you, an impact report, and, on a first gift, an invitation to connect. Checks the Donor System for new gifts.
New suggestions land in the Approval Queue as pending moves. Open the queue to draft and review them.
Approval Queue
Camille: draft in Alyssa's voice
Voice corpus
The real examples of Alyssa writing that Camille and Ali learn from. The more you add, the truer the drafts.
Strati: pipeline and revenue
Buster: engagement trends
Outbox
Approved, ready-to-act content. In this phase a human carries these out; nothing is transmitted automatically.