The three revenue streams
This is where all the study material pays off: how a producer license becomes “Glacis with underwriting capability.” Grounded in the Glacis one-pager and the Law360 article (Heidi Lawson). GROUNDED · Glacis GROUNDED · Law360
The strategy has evolved. Volume 2 rewrites stream 2: Glacis no longer aims to hold the book — a syndicate writes the paper, Glacis stays the independent evidence layer. See the independence thesis.
Glacis in one line
Section titled “Glacis in one line”Glacis is “The Evidence Layer for AI” — cryptographic attestation of governed AI decisions, zero data egress (only hashes cross the trust boundary; it runs in the customer’s environment), declarative policy enforcement at runtime, and native mapping to frameworks (NIST AI RMF, ISO 42001, EU AI Act, HIPAA, SOC 2, Colorado AI Act). GROUNDED · Glacis
A producer license is the legal on-ramp that converts that capability into underwriting — the ability to sell and price risk, not just sell software.
The three streams Heidi outlined
Section titled “The three streams Heidi outlined”| # | Stream | What Glacis sells | Who pays | License/capability prerequisite |
|---|---|---|---|---|
| 1 | Underwriting (SaaS) | The evidence layer to insurers/MGAs so their AI underwriting/claims emit audit-grade, attested evidence (satisfying the NAIC AI Model Bulletin + Evaluation Tool). | Insurers / MGAs | None (it’s software) — but the regulatory tailwind drives demand. |
| 2 | Producer / MGA | A Tech E&O policy for healthcare AI, sold + underwritten on a carrier’s paper via Glacis software, for commission + profit share. | Healthcare-AI vendors (insureds), via the carrier | This is what licensing unlocks: producer license → entity license (DRLP) → appointment → delegated authority. |
| 3 | TPA | Claims error-source attribution (which model/decision caused a loss) → claim-adjusting revenue. | Insurers / self-insured plans | Separate TPA/adjuster authorization (VERIFY per state). |
GROUNDED · Glacis GROUNDED · Law360
How each stream maps to the regulatory hot zones
Section titled “How each stream maps to the regulatory hot zones”Every stream sits inside the highest-scrutiny zone identified in the Law360 article — which is exactly why governance-as-a-feature is the differentiator: GROUNDED · Law360
- Underwriting SaaS → the AI Systems Evaluation Tool checklist becomes a product-requirements spec (governance, model inventory, bias testing, documentation). See 01 evaluation tool.
- Producer / MGA → healthcare is the highest-AI-adoption (92%) and highest-claims-litigation line (Cigna), so a healthcare-AI MGA must be governance-native. See 07 Tech E&O for healthcare AI.
- TPA → claims AI is the hottest litigation zone (State Farm, Cigna); attribution maps directly to explainability/documentation demands. See 02 AI claims litigation.
- Cross-cutting: Glacis-as-software-vendor is itself a “third party” under the NAIC’s broad definition — a future-licensing threat and a moat (Glacis already produces the documentation regulators will demand). See 01 third-party WG.
Why licensing is the keystone
Section titled “Why licensing is the keystone”Streams 1 and 3 can run on software/TPA authority, but stream 2 — the MGA — is the one that turns Glacis into an underwriter, and it is gated entirely on licensing. That’s why this whole academy exists: licensing is Step 1 of the on-ramp. The full sequence is on the next pages.
The honest-sourcing parallel
Section titled “The honest-sourcing parallel”Glacis’s own discipline — bind every claim to what is actually governed/mediated, no overclaim — is the same rule this academy runs on: teach stable concepts, cite the docs, and never fabricate a volatile number. The product and the study aid share a spine. GROUNDED · Glacis
Drill this page →4 bank questions stand behind what you just read. Check it while it’s warm.