These demos show what becomes visible when plans, portfolios, and philanthropic flows are structured as networks rather than documents.
Each one takes a real dataset, applies deterministic extraction with full source traceability, and surfaces the structural patterns — concentration, bridges, latent partnerships, governance silos — that aggregate reporting tends to flatten.
A Maddaus-authored conservation plan, structured as a network. Every node traces back to a page in the source PDF.
Two years of IRS-990 filings from Utah foundations, structured as a network. $165M flows, 8 funders, 761 grantees.
Fifteen foundations, $337M in grantmaking, 254 board members across sixteen boards — and zero interlocks. A governance lens on what advisors to nonprofit boards can actually act on.
Eleven Maine foundation boards, $260M in education giving, and zero board interlocks. What the network lens shows to advisors who work with nonprofit governance.
Two City of Santa Barbara water plans (2020 conservation, 2026 climate adaptation) joined with grant flows from nine California water and climate funders. The first cross-lens demo in the portfolio.
Every node and edge traces back to a source — a PDF page, a 990 filing, a roster entry. Nothing is inferred or invented. That means anyone reading the graph can verify what it claims.
The graphs are curated, not comprehensive. Showing all 1,880 nodes in the Maine network would be a hairball; showing the ~15 that matter is a story. Curation is part of the work.
Structural options, not recommendations. The demos surface what is structurally possible — a pooled fund, a named delivery partner, a governance question worth asking — without claiming to know the political or operational realities of who should actually move.