// TECHNICAL WHITEPAPER

Human Signal Intelligence Protocol — Whitepaper

Complete technical documentation of Foremetric.AI

// TABLE OF CONTENTS

01.Model Collapse: The Synthetic Data Crisis

Modern AI models increasingly train on data they themselves generate. This creates a closed loop of degradation — a phenomenon described by Oxford and Cambridge researchers as Model Collapse. When models consume their own output instead of genuine human signal, they lose the ability to reflect real human behavior.

The problem scales: as synthetic data grows as a proportion of internet content, each new generation of models drifts further from reality. AI literally forgets how real people behave. Research (Shumailov et al., Nature 2024) proves: models progressively and irreversibly degrade across generations when trained on recursively generated synthetic data.

ForeMetric solves this problem at its root. We build infrastructure for collecting, verifying, and monetizing authentic human behavioral data — Human Signal Intelligence. Every data point is cryptographically confirmed as originating from a real person, not a bot or synthetic agent.

"Without live data from real people, the next generation of AI will be trained on a reflection of itself. Model Collapse is not a theory. It is already happening."

02.Protocol Architecture

TEE Protection & Data Sovereignty

All user data is designed to be processed inside a Trusted Execution Environment (TEE) — a hardware-isolated enclave where even the server operator would not be able to access raw data (planned, Phase 2). TEE is designed as a "black box": data goes in, aggregated insights come out, but nobody — including ForeMetric — sees individual records (planned, Phase 2).

Each TEE session is designed to generate a cryptographic attestation confirming the code inside the enclave has not been modified. This proof is designed to be recorded on the TON blockchain, creating an immutable audit trail of data processing (planned, Phase 2).

Proof-of-Behavior: Signal Verification

Proof-of-Behavior (PoB) is a cryptographic verification mechanism confirming behavioral data originates from a real person. PoB analyzes interaction patterns: keystroke rhythm, scroll velocity, navigation pauses, micro-gestures — and generates a unique behavioral fingerprint.

This fingerprint cannot be forged by a bot or script. Each verified session is recorded as a transaction on TON, creating a transparent chain of data authenticity.

🛂 Signal Passport — your digital key to your behavioral profile

Signal Passport is not Passive Signal Earnings in the background. It is designed as a personal cryptographic token that would give third-party services limited and fully user-controlled access to your behavioral profile (planned, Phase 2; not deployed).

How it works:

In the app you would configure (planned, Phase 2):

  • Which behavioral data categories can be used
  • How many queries are allowed (1 to 9,999)
  • Passport validity (24 hours, 7 days, 30 days, or unlimited)

By design, when a service (store, insurance, gift platform, etc.) wants to make a personalized offer, it would scan your Signal Passport QR code and send a request to our API. We would return only the behavioral vectors you permitted. The service would create the most accurate offer based on your real behavior (planned, Phase 2; not deployed).

You are designed to receive $HSIG upon data transfer, whether or not you accept the offer (planned Phase 2).

Real-life use cases for ordinary users:

🎁 Birthday Gift

Imagine: it's your friend's birthday. Under the planned Phase 2 design, you would visit a gift platform and scan their Signal Passport QR code. Your friend would have pre-approved access to "interests" and "hobbies" categories (not deployed).

→ The service is designed to analyze your friend's behavioral profile (planned Phase 2)

→ Suggests 3–5 personalized gifts with 90%+ accuracy

→ Your friend is designed to earn $HSIG for every query to their data (planned, Phase 2; not deployed)

→ No personal data is shared — only aggregated behavioral vectors

🏥 Personalized Insurance

Instead of filling out 40 questions — one QR code. By design, the insurance company would scan your Signal Passport and receive a verified behavioral profile: lifestyle, driving style, physical activity level (planned, Phase 2; not deployed).

→ You get an individual rate instead of "market average"

→ Up to 30% discount for active lifestyle users

→ Planned $HSIG payment for data transfer (Phase 2) — even if you don't buy insurance

→ 30-second process instead of 20 minutes

🥗 Personalized Nutrition & Fitness Recommendations

By design, a fitness app or nutritionist would scan your Signal Passport. Based on your real life rhythm (not a questionnaire, but actual behavior), the system would determine optimal times for workouts, meals, and rest (planned, Phase 2; not deployed).

→ Optimal workout time: when your body is most active

→ Nutrition recommendations based on real habits, not "average" data

→ Personalized recovery and sleep plan

→ You are designed to earn $HSIG for every query to your profile (planned, Phase 2; not deployed)

💼 Career & Job Matching

By design, a recruiting platform would request your Signal Passport to evaluate a fit-score — how well you match a specific position (planned, Phase 2; not deployed). No resume, no first-round interview.

→ Behavioral fit-score (productivity, communication style, habits) without revealing personal data

→ Company gets objective data instead of a subjective interview

→ You are designed to earn $HSIG for every query — even if you don't get the offer (planned, Phase 2; not deployed)

→ You control which data categories are available to the recruiter

Sovereign Data Exchange (SDE)

SDE is designed as a data marketplace where companies would buy aggregated behavioral insights for $HSIG tokens, and users would receive rewards. Users would have full control over which data categories they allow for sale. SDE is designed around an opt-in model: no category would be sold without explicit consent (planned, Phase 2; not deployed).

03.How Users Earn

🔋 Passive Signal Earnings

Just use your phone as normal. During foreground sessions (15–30 min/day) and background activity recognition, the app captures verified behavioral signals with your consent. Battery target: under 2% per day.

The average user generates thousands of verified behavioral micro-signals per session. The longer the app runs, the higher your Sovereignty Score, and the more you earn.

🎯 Missions from Businesses

By design, companies would publish tasks (missions) directly in the app. Examples: "Test our new app and leave feedback," "Take a survey about food preferences," "Rate our website's UX." You would earn $HSIG for every completed mission (planned, Phase 2; not deployed). Missions would always be voluntary — you would choose which ones to complete.

For businesses, this is a direct channel to a real, verified audience. No bots, no fake responses — every participant is confirmed through Proof-of-Behavior.

📊 What Businesses Get (B2B)

Businesses need our data because ForeMetric provides what Google Analytics, Meta, or any other tracker cannot: verified behavioral data from real people, not bots.

// B2B METRICS

→Real behavioral patterns (not bot clicks)
→Verified demographics and audience segments
→Predictive analytics based on real behavior
→Ready-to-use predictive insights for business decisions
→Mission results: UX tests, surveys — fraud-free

By design, companies would pay $HSIG for data access through the Sovereign Data Exchange, and users would decide which data to sell (planned, Phase 2; not deployed). All data is designed to be aggregated and anonymized through TEE (planned, Phase 2). This is designed as a win-win: business gets quality, users get money and control.

🔗Referral System — Passive Income (Planned, Post-TGE)

The ForeMetric referral system is designed to operate on the TON blockchain without cascades or multi-level schemes (planned, post-TGE; not deployed).

By design, when you invite someone via your referral link, a permanent Soulbound connection would be recorded on the blockchain between your wallet and the invited person's device (planned, post-TGE; not deployed).

As long as that device is active in the protocol, you would automatically receive 10% of its income from behavioral signal collection (planned, post-TGE) and every Signal Passport usage. Without any extra action. Forever. Via smart contract.

Activation deferred until wallet verification (wallet_verified=true); intent recorded at first registration, immutable thereafter (per REG-WIRE-016 product decision).

// USE CASE

Under the planned design, a blogger would place their referral link under a video once. If 3,000 people register within a month, the blogger would receive 10% of the income from those devices (from both behavioral signal collection and Signal Passport usage) post-TGE. This is designed as a permanent on-chain asset, but is not deployed.

04.$HSIG Tokenomics

The $HSIG token is the unit of exchange for verified behavioral data in the ForeMetric ecosystem. Total supply: 1,000,000,000 (1 billion) $HSIG. Blockchain: TON (The Open Network).

// KEY FACTS

  • • Total supply: 1,000,000,000 $HSIG — designed to be fixed (planned, not yet deployed).
  • • Emission: zero. The contract is designed with no mint() function (planned, not yet deployed).
  • • Revenue model: 80/10/10 Real Yield Splitter — 80% of B2B protocol revenue → users (via automatic $HSIG buyback on DEX), 10% → referral partners, 10% → protocol treasury.
  • • The splitter ratio is designed to be hardcoded in immutable TVM bytecode at mainnet deployment: no admin key, no governance override.

Users earn $HSIG in two ways: Passive Signal Earnings and completing business missions. Companies pay $HSIG for access to aggregated insights via the COCOON network (planned, Phase 2). This creates a closed-loop economy where value is based on authentic human data.

Dev Fund

The Dev Fund is the protocol development fund, managed exclusively by the Architect and technical contributors. This is not a DAO. This is not community voting. This is an engineering fund managed by those who build the protocol. It is financed by the pre-seed round and public donations.

// FUNDING TARGET

Pre-Seed $250K–$300K: 12–14 months runway, core team expansion (Lead Core Architect, Mobile Systems Engineer), AMD SEV-SNP (Phase 2, Q1 2027) bare-metal deployment & COCOON integration, smart contract audit, B2B pilot program (upcoming) and legal structure.

The Dev Fund accepts public donations in TON. All funds are spent exclusively on development: infrastructure, security audits, engineering hires. Spending transparency is ensured through public reports.

Donation address: UQAnnqbxHVTYv3ANyLw3kIcEBKpZf6WoltHKcXiIE2O-Jl7q

05.Roadmap

Q3–Q4 2026

Telegram Mini App alpha targeting December 2026 (not yet started)

Smart contract on Testnet (33/33 tests passed). GitHub open source. Android APK alpha via Telegram Mini App onboarding hub. Alpha — targeting up to 500 participants (Q4 2026, not yet started). TON Foundation grant application submitted. B2B pipeline pre-qualification begins.

Q1–Q2 2027

Phase 2: TGE & Hardware Attestation

$HSIG Fair Launch on TON (DeDust.io + Ston.fi). AMD SEV-SNP (Phase 2, Q1 2027) bare-metal deployment and COCOON integration. Rust Core + Native App launch. Full B2B Behavioral Queries live. Public team profiles disclosed.

Q3–Q4 2027

Phase 3: Scale & PoB

1D-CNN Proof-of-Behavior classifier (planned Phase 2). Signal Passport discovery pilot (planned Phase 2). 50K MAU target. DEX buyback engine (planned, post-TGE) — 80% of gross B2B receipts designed to reach users. First paying enterprise clients.

2028+

Phase 4: Full Decentralization

Federated Learning (COCOON Secure Aggregation) (Phase 3, 2028+). Kyber-768 Post-Quantum migration. Global scale, wearables integration, EU and SEA market entry.

06.Conclusion

ForeMetric is not another data collection project. It is an infrastructure shift: from a model where corporations extract data for free, to one where every person owns and monetizes their behavioral signal.

We solve a fundamental AI industry problem — Model Collapse — by providing the only scalable source of verified human data. TEE protection is designed to guarantee sovereignty (planned, Phase 2), Proof-of-Behavior to guarantee authenticity, and the $HSIG token to enable fair exchange economics (not deployed).

"We don't sell your data. We create a market where you decide how much your living signal is worth — and get real money for it."

What is Model Collapse?

Model Collapse is a phenomenon where AI models degrade when trained on synthetic data generated by other models. Oxford research (Shumailov et al., Nature 2024) proves: models progressively and irreversibly degrade across generations when trained on recursively generated synthetic data. ForeMetric solves this by providing verified human behavioral data through the Human Signal Intelligence protocol.

How does ForeMetric protect privacy?

On-device, data is designed to be processed inside ARM TrustZone / Apple Secure Enclave so raw behavioral data would not leave the phone (planned, Phase 2 native app). Server-side (Phase 1): disclosed trusted-operator key custody (LUKS + AWS KMS). Server-side (Phase 2): full AMD SEV-SNP + COCOON TDX hardware enclaves are planned, where even the host operator would not be able to access memory. Each session is designed to generate a cryptographic attestation recorded on the TON blockchain (planned, Phase 2).

What is Proof-of-Behavior?

Proof-of-Behavior (PoB) is a cryptographic verification mechanism that confirms behavioral data comes from a real human, not a bot. It analyzes interaction patterns — keystroke rhythm, scroll velocity, micro-gestures — demonstrating 93–97% target bot-filtering accuracy during simulated validation on synthetic behavioral datasets. Production accuracy thresholds will be formally calibrated during live Alpha cohort deployment.

How do users earn $HSIG tokens?

Users would install the ForeMetric app, which is designed to passively collect anonymized behavioral data with consent (planned, Phase 2). 80% of gross B2B receipts is designed to buy back $HSIG on TON DEX for active contributors (planned, post-TGE). Additional income is designed to come from the referral system (10%, planned, post-TGE) and Signal Passport — a planned configurable access token to your behavioral profile (Phase 2; not deployed).

What is Signal Passport?

Signal Passport is a planned C2C/B2B personalization experiment (discovery pilot). Access controls, payments and revocation described here are design, not deployed capability. Planned design: query limits (1 to 9,999), data sectors, expiry, $HSIG payment per verified query and on-chain revocation.

Download Whitepaper (PDF)

Public edition — October 2026

Extended edition with evidence registry — available to qualified investors upon request via foremetric.ai