This is a working demo of Surface, a research intelligence platform built by Comprendo. It demonstrates how AI-assisted research can map a complex landscape — identifying the key players, funding flows, power dynamics, and strategic opportunities that would normally take a team months to assemble.

The subject of this demo is K-12 data science education in the United States — a nascent field with ~70,000 enrolled students, $20-30M in annual philanthropic investment, and a small but influential network of funders, curriculum providers, and field-builders shaping its direction.

What's in This Demo

Each tab represents a different lens on the same underlying research:

Analysis Written intelligence

A six-part landscape analysis covering field context, the funder landscape, grantee and organization profiles, the work itself (curricula, standards, the Algebra 2 debate), a network and power map, and a strategic assessment with five identified white spaces for investment. This is the kind of deliverable a consulting engagement produces — except assembled in days rather than months.

Network Interactive visualization

A force-directed graph of the entities and relationships in the landscape. Funders, field-builders, curriculum providers, and their connections — rendered as an interactive map you can filter, zoom, and click through. Reveals structural patterns like co-funding clusters and the central role of coordinating bodies that are invisible in written reports.

Entities Structured profiles

Detailed cards for every entity in the landscape — funders, the central coordinating body (DS4E), curriculum providers, field-builders, and key people. Each profile includes funding details, reach, strategic role, and connections to other entities. A quick-reference layer for anyone navigating the space.

Signals Monitoring feed

Growth signals, fragility signals, and open questions — the kind of ongoing intelligence that keeps a picture current. Tracks developments like state policy momentum, funder exits, financial distress at key organizations, and unresolved strategic questions. Designed to be updated as the landscape evolves.

Ask Conversational query

A natural-language interface to query the research knowledge base. Ask questions like "Who funds CourseKata?" or "What states have adopted data science standards?" and get sourced answers drawn from the underlying research. Turns a static report into a living, queryable resource.

How It Works

Surface combines AI-assisted research with human editorial judgment. The process starts with broad source collection — public filings, grant databases, organization websites, policy documents, news coverage — then uses AI to extract entities, relationships, and signals from those sources. A human researcher reviews, validates, and synthesizes the output into the structured views above.

The result is not a chatbot summary. It's a structured intelligence product with specific claims, named sources, and an explicit point of view. The AI accelerates the grunt work — reading hundreds of documents, cross-referencing funding data, identifying patterns — while the human ensures accuracy, context, and strategic framing.

Why This Approach

Days, not months

A landscape analysis of this depth — mapping funders, grantees, power dynamics, and strategic opportunities — typically takes a research team 8-12 weeks. This demo was assembled in a fraction of that time. The AI handles volume; the human handles judgment.

Structure from the start

Research usually lives in documents and slide decks. Surface produces structured, queryable data from day one — entity profiles, tagged relationships, classified signals. This means the research is immediately usable, not buried in a 50-page PDF.

A living picture

Traditional landscape analyses are snapshots — accurate on delivery day, stale within months. Surface's signal monitoring and queryable knowledge base are designed to stay current as the landscape evolves, turning a one-time deliverable into an ongoing asset.

Bootstraps your expertise

A new funder entering this space would need months to build the mental model that this platform provides in an afternoon. Surface doesn't replace domain expertise — it accelerates the journey from "I don't know what I don't know" to "I know the right questions to ask."

What This Demo Reveals

Beyond demonstrating the platform, this particular landscape analysis surfaced findings that don't exist elsewhere in public:

No one maps who funds K-12 data science education. DS4E's own 79-page State of the Field report does not name a single funder. This analysis is, to our knowledge, the first public attempt to map the funding architecture of this field.
The top 3 private funders dominate $20-30M in annual investment. A new funder committing $1-3M would immediately become a top-five player — the field is small enough that strategic investment at this scale can shape its direction.
Key infrastructure is fragile. ExcelinEd, the field's primary state policy organization, shows a 74% revenue decline. CZI and Schmidt Futures have reduced or exited their commitments. The funder base is narrowing at a moment of field growth.

Interested?

Surface can be applied to any landscape where you need to understand the players, the money, the power dynamics, and the opportunities — education, climate, health, policy, technology. If you're a funder, foundation, or organization trying to navigate a complex space, get in touch.