What is the total addressable market for ONTO?

$50B+ across regulated industries requiring verifiable AI quality: Healthcare AI ($45B by 2030), Financial AI ($44B), GovTech ($40B), Defense AI ($24B), Legal AI ($3.3B). Every sector adopting AI regulation needs an enforcement instrument — ONTO is the only one that exists.

What are ONTO's competitive moats?

Five moats: (1) 104-byte proof chain — uncopyable cryptographic architecture, (2) R7+ modules — primary long-term moat beyond commoditized R1–R6, (3) GOLD Core — 169 files, 900K tokens of cross-science epistemic architecture, (4) 20 years of foundational research adapted for AI, (5) First-mover in a category that becomes mandatory with AI regulation.

Does ONTO have competitors?

ONTO created a new category: epistemic discipline for AI at inference time. Benchmarks (MMLU, HELM) test what models know. Guardrails restrict what models say. ONTO disciplines how models communicate — cite, quantify, disclose, admit unknowns. No other system provides cryptographic epistemic proof per response. Zero direct competitors.

What is ONTO's revenue model?

Four tiers: OPEN ($0, 10 calls/day), STANDARD ($2,500/month, enterprise), PROVIDER ($250,000/year, AI companies), WHITE-LABEL ($500,000/year, own branding). Government revenue: national mandate certification fees. Pre-revenue as of April 2026, first pilots targeting Q2–Q3 2026.

How does ONTO compare to AMI?

AMI: 6 modules, $1.03B seed funding (March 2026), zero shipped products, first product expected 2027. ONTO: 18 R-modules (each deeper than AMI's — 1R equals 6 AMI by quality), production-ready product, 22+ models tested, published results. ONTO is further with less capital.

Yes. Pre-revenue as of April 2026. Product 1 (Regulator) is production-ready with live API. First enterprise pilots targeting Q2–Q3 2026. Government outreach across 9 countries in progress. Worst-case 18-month revenue projection: $1.3–4.9M.

What does ONTO need from investors?

Strategic partner — government fund, sovereign wealth, or institutional investor who understands regulatory infrastructure. Not just capital: access to regulators, credibility in target markets, advisory support for scaling a standards body.

The structured epistemic knowledge architecture powering both ONTO products. 169 files across 7 scientific domains, approximately 900K tokens. Injected at inference time — models are never modified. Not a prompt, not fine-tuning. A curriculum that teaches AI seven capabilities: cite sources, quantify confidence, grade evidence, present counterarguments, calibrate confidence, state falsifiability, admit unknowns.

What evidence exists that ONTO works?

22+ models tested across 12 published reports. Key results: composite quality 10× improvement, unknown recognition 26× improvement, source citation from 0 to 3+ per response. Scoring is deterministic — same input, same score, every time. All data published at github.com/nickarstrong/onto-research.

What is the current team?

Solo founder with 20 years of cross-science research adapted for AI. Advisory board forming (minimum 3 people). Hiring priorities: CTO and GTM. Legal entity in structuring — jurisdiction TBD.

What is ONTO's origin story?

20 years of cross-science research — epistemology, cognitive science, clinical methodology, statistical inference, information theory, formal logic, decision science. Not invented for AI — discovered, calculated, coded, cross-verified, and adapted. The research predates the AI boom. ONTO translated existing scientific discipline into a format any AI can follow.

EU AI Act enforcement begins 2025–2026. Every regulated industry needs verifiable AI quality — the enforcement instrument didn't exist until ONTO. First-mover advantage in a category that becomes mandatory. 9-country pipeline with regulatory demand already articulated. Technology proven with published data. The window is open now.

ONTO — Epistemic Discipline Layer for AI

Q: What is ONTO and why should investors care?

ONTO is the epistemic measurement standard for AI — three deployments of one standard. Regulator: grades every AI A–F with Ed25519 cryptographic proof chain. Agent: BYOK measurement app for business and individuals. Human AI: R1–R18 discipline layer for providers and robotics, retrofit or born-disciplined. $50B+ total addressable market in regulated AI. 22+ models tested with published results. Deterministic, inference-time, per-response.

ONTO — The Discipline Standard for AI Quality · Investor Brief

ONTO is the epistemic measurement standard for AI. Three deployments of one standard: Regulator (A–F grading with Ed25519 cryptographic proof chain), Agent (BYOK measurement app for business and individuals), Human AI (R1–R18 discipline layer for providers and robotics — retrofit or born-disciplined). Total addressable market exceeds $50B across regulated industries. Deterministic scoring, inference-time, measurable per response.

Market opportunity

AI regulation is becoming mandatory worldwide — EU AI Act, Singapore AI Governance, national AI strategies across 9 target countries. Every regulated industry needs verifiable AI quality: Healthcare AI ($45B by 2030), Financial AI ($44B), GovTech ($40B), Defense AI ($24B), Legal AI ($3.3B). The enforcement instrument for AI regulation did not exist until ONTO. Governments pass laws but lack the tool to enforce them. ONTO fills this gap.

Technology and moats

GOLD Core: 169 files, approximately 900K tokens of structured epistemic architecture across 7 scientific domains. Injected at inference time — no model modification. Deterministic dual-layer scoring: Python linguistic analysis plus Rust statistical analysis, 993 lines, zero AI in evaluation. 104-byte proof chain: timestamp plus SHA-256 plus Ed25519 signature. Five moats: uncopyable proof architecture, R7+ deep modules, 20 years of foundational cross-science research, first-mover in mandatory category, GOLD Core depth.

Evidence

22+ models tested across 12 published reports. Composite quality improvement: 10×. Unknown recognition: 26× improvement. Source citation: 0 to 3+ per response. Calibration: 0 to 1.0. All results deterministic and reproducible. Published at github.com/nickarstrong/onto-research.

Competitive landscape

AMI: $1.03B seed (March 2026), 6 modules, zero shipped products, first product expected 2027. ONTO: 18 R-modules (each deeper — 1R equals 6 AMI by quality), production-ready, published evidence, fraction of the capital. Benchmarks (MMLU, HELM) measure knowledge. Guardrails restrict output. ONTO disciplines how AI communicates — a fundamentally different and complementary category.

Revenue model

Four pricing tiers from free to $500K/year. Enterprise: $2,500/month. AI providers: $250K/year. White-label: $500K/year. Government: national mandate certification fees. Pre-revenue as of April 2026. First pilots Q2–Q3 2026. Worst-case 18-month projection: $1.3–4.9M revenue against $1.5–2M burn.

Government strategy

9 countries targeted simultaneously: Uzbekistan, UAE, Turkey, Singapore, Kazakhstan, Saudi Arabia, Japan, Germany, USA. Each shown the other 8 as competition. First to sign gets exclusive terms. ONTO provides enforcement instrument: AI grading, certification, and revenue generation for governments.

Current status

Product 1 (Regulator): production-ready, API live, scoring deterministic. Product 2 (Human AI): protocol 100%, SaaS in development. Solo founder, advisory board forming. Legal entity structuring in progress. Pre-revenue. Government outreach April 2026.

The ask

Strategic partner — government fund, sovereign wealth, or institutional investor with access to regulators and credibility in target markets. Not just capital: advisory support for scaling a standards body into a global institution.

Situation	Regular AI	Restricted (RLHF)	Human AI (R1-R18)
Medication	Recommends. No source.	Consult your doctor.	Patikorn 2022, n=410, CI 70%
Doesn't know	Fabricates	Complex topic.	3 scenarios + probabilities
Contradiction	Picks one side	Different opinions.	Both + evidence weight
Self-aware	No	No	R16: monitors own quality

Model	Without GOLD	With GOLD	Δ
GPT	6.5/C	9.9/A	+52%
Grok	6.7/C	9.3/A	+39%
DeepSeek	7.3/B	8.9/A	+22%
Gemini	4.1/D	7.8/B+	+90%
Claude	6.5/C	9.9/A	+52%

#	Module	Meta AMI	ONTO Human AI
1	World Model	V-JEPA, research only	GOLD: 7 domains, 3 levels
2	Perception	Partially	Scoring: 993 lines, EM1-EM5
3	Critic	Not built	Dual-layer: Python + Rust
4	Actor	Not built	Proxy + Agent, production
5	Memory	Not built	169 files + Ed25519
6	Configurator	Not built	Router + Kernel R1-R7
7	Epistemic Layer	NOT IN PLAN	R2 + R7 — deployed
8	Disciplined Creativity	Not in plan	R8 — scenarios + probabilities
9	Domain Specialization	Not in plan	R9 — medicine, law, finance
10	Multimodal Verification	Not in plan	R10 — cross-check channels
11	Causal Reasoning	Not in plan	R11 — correlation ≠ causation
12	Temporal Calibration	Not in plan	R12 — 2024 ≠ 2020
13	Adversarial Resilience	Not in plan	R13 — jailbreak protection
14	Epistemic Audit	Not in plan	R14 — decision log
15	Collaborative Verification	Not in plan	R15 — models check each other
16	Epistemic Self-Awareness	Not in plan	R16 — 3 models demonstrated

Stream	Model	Who pays
Provider Certification	$250K/year	OpenAI, Anthropic, xAI, Google, DeepSeek
State License	$500K–2M/yr	Dashboard + enforcement + proof chain for regulators
Domain Licenses	$100K–1M/yr	Hospitals, banks, law firms
Human AI Protocol	Partnership	Strategic partners — R8–R18 cognitive architecture

	2026	2027	2028	2029
🔴 Pessimistic	$50K	$1-5M	$5-10M	$10-15M
🟢 Base	$100K	$5-8M	$15-20M	$35-50M
🟣 Optimistic	$200K	$10-15M	$30-40M	$70-100M

ONTO — The Discipline Standard for AI Quality · Investor Brief

Market opportunity

Technology and moats

Evidence

Competitive landscape

Revenue model

Government strategy

Current status

The ask

One standard. Three deployments.
Every AI disciplined.

DISCIPLINE (R1-R7) — AI stops fabricating

COGNITION (R8-R18) — AI starts thinking

Same model. Same question. One layer. 0.12/F → 8.85/A.

We built what they couldn't.

Nobody teaches AI where the edge is.
Everyone forbids it from crossing.

Grok banned entirely

€17B on compliance

Tied hands in the OR

What happens after you say yes

	2026	2027	2028	2029
Revenue	$100K	$5-8M	$15-20M	$35-50M
Costs	$200K	$1.5M	$5M	$10M
Profit	-$100K	$4-6.5M	$12-15M	$28-40M
Margin	pilot	~75%	~78%	~80%

Customer	CAC	LTV	LTV/CAC
Government	$45K	$5M+	110×
Provider (direct)	$20K	$1M	50×
Provider (mandate)	$3K	$1M	333×
Institution	$10K	$900K+	90×

Tier	Price	Access	For
OPEN	$0	10 calls/day · Full GOLD	Testing
STANDARD	$2,500/mo	1,000 calls/day	Enterprises
PROVIDER	$250K/yr	Unlimited + SSE	AI providers
WHITE-LABEL	$500K/yr	Unlimited + SSE + branding	Gov partnerships

#	Country	Entry Point	Killer Fact	Status
1	🇺🇿 Uzbekistan	Personal contact	First in CIS. $1B AI budget.	Docs ready
2	🇦🇪 UAE	MBZUAI	$169B + Falcon + MBZUAI	Docs ready
3	🇹🇷 Turkey	Embassy Tashkent	Grok ban. EU bridge.	Docs ready
4	🇸🇬 Singapore	Direct	AI Verify + ONTO = full stack	Docs ready
5	🇰🇷 S. Korea	Inha Univ. Tashkent	K-AI Quality brand	Docs ready
6	🇸🇦 Saudi	Direct	Vision 2030 + $100B+ invested	Docs ready
7	🇯🇵 Japan	Direct	AI Basic Act pending	Docs ready
8	🇩🇪 Germany	Direct	EU AI Act homeland	Docs ready
9	🇺🇸 USA	Last	NIST AI RMF compatible	Docs ready

Country	Law / Framework	Pain	ONTO trigger
🇹🇷 Turkey	AI Bill. Grok banned Jul 2025.	First AI ban in history.	Grade, don't ban.
🇪🇺 EU	AI Act 2025-2027. €35M fines.	€17B compliance. No grading.	Automated grading + proof.
🇺🇿 Uzbekistan	AI Law 2026. $1B budget.	100+ AI projects, zero QA.	First in CIS. Dashboard.
🇦🇪 UAE	AI Gov Framework. Strategy 2031.	Framework, no enforcement.	Build + research + certify.
🇸🇬 Singapore	Model AI Gov. AI Verify.	Voluntary. No verification.	AI Verify + ONTO = full stack.
🇰🇷 S. Korea	AI Basic Act, in force.	Companies unclear on instrument.	K-AI Quality brand.
🇸🇦 Saudi	AI Strategy 2030. SDAIA.	Billions invested, no QA.	ROI discipline.
🇯🇵 Japan	AI Guidelines. Act pending.	Guidelines without teeth.	Enforceable compliance.

ONTO — The Discipline Standard for AI Quality · Investor Brief

Market opportunity

Technology and moats

Evidence

Competitive landscape

Revenue model

Government strategy

Current status

The ask

One standard. Three deployments.Every AI disciplined.

DISCIPLINE (R1-R7) — AI stops fabricating

COGNITION (R8-R18) — AI starts thinking

Same model. Same question. One layer. 0.12/F → 8.85/A.

We built what they couldn't.

Nobody teaches AI where the edge is.Everyone forbids it from crossing.

Grok banned entirely

€17B on compliance

Tied hands in the OR

What happens after you say yes

One standard. Three deployments.
Every AI disciplined.

Nobody teaches AI where the edge is.
Everyone forbids it from crossing.