Replate Whitepaper

Onchain proof of healthy living. A receipt intelligence and rewards protocol built for health-focused users in the Base and Farcaster ecosystems.

Version 2 · Living product document

This document separates what is Live, what is Planned, and what represents the broader Strategy.

LIVE

1. Current product

Live

A user uploads or captures a grocery receipt and provides household context.

Google Cloud Vision extracts the receipt text. Replate removes totals, payment lines, and recognized non-food items, then normalizes and classifies products as healthy, unhealthy, or neutral. Optional Open Food Facts enrichment may support analysis.

The user receives a Health Score and Nutrition Score, reviews the result, signs an EIP-712 message, and submits the verification transaction from the connected wallet.

The current frontend therefore preserves a user-signed Base transaction as the primary verification path.

A validator-relayed gasless path exists in the backend but is not the default frontend flow.

2. Scoring and onchain verification

Live

The Health Score gives full weight to recognized healthy products, partial weight to neutral products, and zero weight to unhealthy products.

The Nutrition Score compares fruit-and-vegetable grams with an average reference target of approximately 300g per person per day, adjusted by household size and the period covered.

The blockchain stores only aggregate information such as:

• Health Score
• Nutrition Score
• item counts
• fruit-and-vegetable grams
• household size
• days covered
• points
• timestamp
• receipt hash

Raw images, OCR text, product names, store names, and detailed AI recommendations are not written to the smart contract.

Replate scoring is informational feedback, not medical advice.

3. XP, streaks, and badges

Live

Verified receipts can produce XP and update weekly progress.

Daily check-ins use a separate XP and streak mechanic.

A Replate Badge NFT can be minted when defined health and nutrition thresholds are reached.

These mechanics are designed to reward consistency rather than perfection.

4. Privacy and current data handling

Live

Replate does not intend to retain raw receipt images or full OCR text longer than necessary for product operation.

OCR, normalization, and classification may be inaccurate, particularly for blurry, damaged, multilingual, or unusual receipts.

Current onchain data is pseudonymous rather than fully anonymous because wallet addresses and aggregate activity remain publicly visible on Base.

The current product is designed so that detailed receipt content stays offchain while only aggregate proof data is written to the smart contract.

PLANNED

5. Replate Intelligence data layer

Planned

Replate plans to build a PostgreSQL-based intelligence layer alongside the existing Base contracts.

The proposed pipeline is:

Raw Receipt → Normalized Data → Derived Features → Behavioral History → Replate Intelligence

Potential intelligence includes:

• basket diversity
• processed-food ratio
• protein diversity
• fruit-and-vegetable quality
• household coverage
• 30 / 60 / 90-day trends
• historical improvement
• shopping consistency
• personalized recommendations
• aggregate consumer signals
• permissioned reputation

Blockchain remains the verification and reward layer.

PostgreSQL becomes the normalization, historical analysis, aggregation, and intelligence layer.

6. Premium consumer model

Planned

The basic Replate experience is intended to remain accessible without charging for every receipt.

The preferred consumer flow is:

Receipt Upload → Basic Analysis → User-Signed Base Verification → Optional Advanced Replate Intelligence

After verification, a user may purchase a detailed intelligence report through an x402-compatible USDC payment.

Replate may also introduce Replate Intelligence+:

5 USDC · 30 days

Potential benefits include:

• advanced receipt analysis
• personalized recommendations
• historical trends
• advanced weekly reports
• premium profile status
• premium quests
• early feature access
• selected seasonal benefits
• proposed 1.5x XP multiplier on eligible verified activity

Subscription status would be managed by Replate while x402 acts as the payment rail.

7. x402, MCP, and AI infrastructure

Planned

Replate plans to use x402 for both consumer premium services and machine-to-machine payments.

Consumer model:

Verified User → x402 Payment → Advanced Replate Intelligence

Infrastructure model:

AI Agent / Web3 Application → Replate API or MCP Tool → x402 Payment → Replate Intelligence

Potential paid services include:

• Basket Intelligence
• Nutrition Intelligence
• Behavioral Trend Analysis
• Recommendation Intelligence
• Aggregated Consumer Insights
• Permissioned Reputation Signals

MCP-compatible tools may allow AI agents to access these services directly.

x402 is the payment rail. Replate Intelligence is the product.

8. Founder NFT and sponsored ecosystem

Planned

Replate plans to explore a limited Founder NFT collection:

Supply: 444 Mint Price: 5 USDC

Potential utility includes:

• Founder status
• profile badges
• early access
• Founder-only quests
• selected seasonal benefits
• community privileges
• possible Intelligence+ benefits

Founder NFTs are intended as utility and membership assets rather than guaranteed revenue-sharing products.

Replate may also work with AI companies, Web3 protocols, wallets, Base ecosystem projects, wellness companies, and consumer brands on sponsored quests, seasons, reward pools, or infrastructure campaigns.

Sponsors are an additional growth layer, not the core revenue dependency.

9. Planned roadmap

Planned

Near-term priorities include:

• better retailer parsing
• stronger product normalization
• OCR and classification confidence
• duplicate and fraud detection
• PostgreSQL intelligence architecture
• historical behavioral tracking
• Advanced Replate Intelligence
• x402 premium consumer payments
• Intelligence+ membership
• B2B Intelligence APIs
• MCP-compatible AI tools
• Founder NFT utility
• sponsored ecosystem experiments

A native token and mass-market wallet-abstracted experience remain conditional on proven product-market fit and sustainable usage.

STRATEGY

10. Abstract

Strategy

Replate combines OCR, product classification, household context, AI analysis, and Base blockchain verification to turn grocery receipts into health feedback, verifiable progress, and privacy-conscious behavioral intelligence.

Today, Replate focuses on receipt analysis, Health Score, Nutrition Score, XP, streaks, badges, weekly progress, and user-signed onchain verification.

The long-term vision is to build Replate Intelligence: a structured intelligence layer that can serve users, AI agents, and Web3 applications through premium consumer features, x402-powered APIs, and MCP-compatible tools.

11. Mission and audience

Strategy

Replate helps users understand and improve their grocery-shopping behavior while creating a transparent, verifiable record of progress.

Phase 1 focuses primarily on crypto-native users in the Base and Farcaster ecosystems interested in health, fitness, community, AI, and onchain identity.

The initial objective is to prove repeated receipt usage, retention, useful scoring, reliable data normalization, and sustained onchain participation.

Long term, Replate may also serve AI agents, developers, wellness platforms, and Web3 applications that need structured grocery and behavioral intelligence.

12. The Problem

Strategy

Every year, billions of grocery receipts are discarded after serving a single purpose: proving that a purchase happened.

Yet these receipts contain valuable signals about real-world consumer behavior, including what people buy, how shopping patterns change over time, and how those choices may relate to nutrition, household habits, and healthier living.

Today, much of this information is lost, fragmented, or locked inside closed systems.

Replate's goal is to capture this otherwise wasted data, process it responsibly, and transform it into structured, privacy-conscious intelligence.

The Replate model is:

Discarded Receipt Data → Structured Grocery Data → Verified Behavioral Signals → Replate Intelligence → x402-Powered AI & Web3 Services

Instead of treating receipts as disposable records, Replate aims to turn them into useful intelligence that can circulate within the blockchain and AI ecosystem.

Through Replate Intelligence, derived and aggregated insights may become accessible to AI agents, Web3 applications, and developers through x402-powered services, while raw personal receipt data remains protected.

The objective is to create a healthier data economy where value is not extracted from users and lost inside closed platforms.

Instead:

Users contribute real-world activity. Replate transforms that activity into useful intelligence. AI and Web3 applications gain access to valuable structured signals. Part of the economic value generated can return to the Replate ecosystem and its users.

Replate therefore addresses two connected problems:

valuable real-world receipt data is currently wasted, and users rarely participate in the value created from their own activity.

The long-term goal is to turn this discarded information into a useful, privacy-conscious intelligence layer that benefits users, developers, AI systems, and the broader Web3 ecosystem.

13. Privacy and data principles

Strategy

Replate separates data into three broad classes:

Private Data — raw receipts, OCR text, and exact user purchase history.

Permissioned Data — user-level trends, reputation, or personalized intelligence explicitly authorized by the user.

Aggregated Intelligence — anonymized or sufficiently aggregated behavioral and statistical signals.

The guiding principle is:

Replate monetizes intelligence, not personal receipt data.

14. Economic model

Strategy

Replate is designed around three participants.

Users

Users contribute verified real-world activity and receive health feedback, XP, streaks, badges, onchain progress, optional premium intelligence, and potential future rewards.

AI and Web3 applications

External applications may pay for structured basket intelligence, nutrition signals, behavioral trends, recommendations, aggregate intelligence, or permissioned reputation.

Replate

Potential revenue sources include:

• Intelligence+ subscriptions
• pay-per-use premium analysis
• x402 B2B APIs
• MCP / AI agent usage
• sponsored programs
• selected premium features

A portion of future infrastructure revenue may support ecosystem growth and recurring reward pools.

The legacy contract PAID receipt mechanism is disabled; receipt submissions are free and the x402-powered intelligence endpoint is the separate paid product.

15. Risks and open questions

Strategy

The quality of Replate Intelligence depends on accurate OCR, strong normalization, reliable classification, fraud resistance, sufficient user activity, historical depth, and appropriate privacy controls.

Fake receipts, duplicate submissions, and Sybil activity may weaken both rewards and intelligence quality.

AI companies will only pay for Replate services if the intelligence provides meaningful value that is difficult or expensive to reproduce independently.

x402 enables payment but does not create demand by itself.

Privacy, user consent, data licensing, subscriptions, NFTs, rewards, sponsored campaigns, and future token plans require appropriate legal and compliance review before expansion.

16. Vision

Strategy

Replate begins with a grocery receipt.

Its long-term evolution is:

Receipt Analyzer → Health & Nutrition Tracker → Onchain Healthy-Living Proof → Verified Behavioral Data Layer → Replate Intelligence → x402 AI Services → MCP Agent Infrastructure → Real-World Intelligence Network

The goal is to create a system where users understand and verify their progress, premium users receive deeper intelligence, AI systems gain access to useful privacy-conscious behavioral signals, and Replate generates sustainable infrastructure revenue.

Real behavior. Verified intelligence. Machine-to-machine commerce.

Shop Smart. Eat All. Replate.