Three entities.
One mission.
Lighthouse is structured as a commercial company, a nonprofit foundation, and an open protocol — not because it is complicated, but because child protection requires each of those things, and none of them can do the others' jobs.
TwinFlame Group
Parent company
Lighthouse, Inc.
Delaware C-Corp · Commercial SaaS
Lighthouse Foundation
501(c)(3) · Grants · Free tiers
Lighthouse Protocol
Open Standard · Phase 3 · 2028+
Lighthouse, Inc.
Builds and operates the commercial SaaS suite. Raises venture capital. Revenue from Sentinel, FrontlineOS (employer tier), Hosts OS, Beacon OS, Shield OS, and Shield OS enterprise contracts.
Lighthouse Foundation
Operates free and subsidized tiers. Owns Victim ID Workbench, Restore OS, and nonprofit deployments of FrontlineOS and Youth Risk Navigator. Governed by an independent board. Captures foundation grants and federal funding.
Lighthouse Protocol
Phase 3 open standard. Cross-platform predator intelligence sharing modeled on GIFCT governance. Consortium-governed by member platforms. Requires Phase 1 and 2 credibility to launch.
Lighthouse, Inc.
Delaware C-Corp. Subsidiary of TwinFlame Investments, LLC. Builds and operates the commercial SaaS platforms: Sentinel for Parents, FrontlineOS (employer tier), Hosts OS, Beacon OS, Shield OS.
Raises venture capital from investors who understand that mission-driven infrastructure and commercial returns are not opposites. The companies that built SWIFT, Cloudflare, and Stripe proved that. We intend to prove it here.
Revenue from commercial SaaS subscriptions, enterprise contracts, and government procurement (where consistent with civil-liberties guardrails).
Lighthouse Foundation
501(c)(3) — filing in process. Independent governance board. Operates the platforms where nonprofit funding models fit better than commercial ones: Victim ID Workbench, Restore OS, FrontlineOS (worker free tier), Youth Risk Navigator (nonprofit deployments).
Captures foundation grants, federal grants, and philanthropic funding that would be unavailable to a commercial entity. Publishes an annual independently-audited public report on platform impact and governance. Hires a Chief Privacy Officer before signing any federal LOI.
The Foundation and Inc. are related but governed separately. The Foundation does not report to the Inc. board.
Lighthouse Protocol
Phase 3 — 2028+. An open standard for cross-platform predator intelligence sharing. Consortium-governed, modeled on the Global Internet Forum to Counter Terrorism (GIFCT) architecture adapted for child protection.
A predator banned from Discord today re-registers on Roblox tomorrow. The Protocol is the infrastructure that makes platform-hopping a dead end. Platforms contribute hashed behavioral fingerprints into a consortium-governed encrypted vault. Match signals — not underlying data — are returned on new account creation.
The Protocol requires Phase 1 and Phase 2 credibility to launch. We build trust before we ask platforms to trust us with their users.
Five refusals that define
what we will never become.
We refuse to build surveillance tools disguised as safety tools.
Sentinel for Parents does not give parents access to every message. FrontlineOS does not record workers without consent. Beacon OS does not expose child identifiers to unauthorized parties. Protection and surveillance are not the same thing. We will not conflate them.
We refuse to route any flag to vigilante groups, social media, or unvetted individuals.
Every flag fired by a Lighthouse platform routes to a credentialed professional — a case worker, a trained investigator, a vetted analyst. Never to a self-organized group. Never to a public tip page. The moment detection becomes mob justice, it stops being protection.
We refuse to be weaponized for any political agenda.
Child safety is not partisan. Lighthouse platforms will never be deployed for immigration enforcement, political opposition research, or any purpose outside the specific protection of children. If a government contract requires otherwise, we will decline it.
We refuse to let AI act without a human in the loop.
Every flag that could affect a child — or a suspect — requires a trained human being to review it before any action is taken. The system is never permitted to act alone. AI is the intelligence layer. Humans are the decision layer.
We refuse to stop at detection.
Lighthouse exists at two moments: the moment before the harm, and the moment after the disappearance. It does not end at the tip or the report. The platform follows the arc of what happens to a child — and to a survivor — all the way through. Rescue without restoration is not enough.