The Protocol: How It Works
A transparent, decentralized system for food health truth. Three core components working together.
The Food Truth Oracle
4-Layer Data Validation
A cryptographic system that guarantees food data honesty. Submissions are bonded. Disputes are resolved transparently. Truth is verifiable.
The Co-Nutritionist
Multi-Agent AI Tournament
Multiple AI agents analyze each food across 6 metabolic dimensions. They debate, they reason, they cite science. Consensus emerges from diversity.
Decentralized Governance
Community Control
Token holders vote on protocol evolution. Scientists guide the framework. Communities resolve disputes. Power is distributed.
Layer 1: Submission & Bonding
When a food producer submits their product for assessment, they post a cryptographic bond. This stake is at risk if they submit false data.
- ✓Producer submits: nutritional data, ingredients, sourcing, manufacturing details
- ✓Bond is locked: economic disincentive for dishonesty
- ✓Data is timestamped and hashed: immutable record on chain
- ✓Submission enters the validation pipeline
Layer 2: Automated Verification
Automated checks catch obvious errors and inconsistencies.
- ✓Nutrient calculations: do the numbers add up?
- ✓Ingredient cross-checks: does this ingredient exist?
- ✓Manufacturing validation: is this facility real?
- ✓Toxin screening: does this need third-party lab analysis?
Layer 3: Community Challenges
Community members can challenge submissions. If evidence supports the challenge, the dishonest submitter loses their bond.
- ✓Challenge period: window for disputes
- ✓Evidence required: specific allegations with citations
- ✓Community vote: PFF token holders judge validity
- ✓Consequences: bond slash on loss, reward on win
Layer 4: Verification & Certification
Data that survives challenges is cryptographically verified and ready for AI analysis.
- ✓Data is locked: immutable record of what was submitted
- ✓Verification badge: cryptographic proof of validation
- ✓Timestamp: when data was verified
- ✓Ready for scoring: AI system can now analyze
The 6 Dimensions of Metabolic Health
Every food is scored across these dimensions. The Co-Nutritionist AI tournament analyzes each independently, then synthesizes a comprehensive picture.
Blood Sugar & Insulin Impact
Glycemic index, insulin response, glucose variability
Gut Biome Effects
Prebiotic/probiotic effects, SCFA production, microbiome diversity
Toxicity Profile
Pesticide residues, heavy metals, mold toxins, inflammatory compounds
Nutrient Bioavailability
Absorption rates, chelators, nutrient density, mineral balance
Protein Quality
Amino acid profile, PDCAAS score, digestibility, muscle synthesis potential
Universal Metabolic Health Score
Synthesis across all dimensions. How does this food affect your overall health?
Complete Transparency
Every score comes with a complete reasoning chain showing exactly how it was calculated.
- ▸Input data: exactly what was submitted, when, and by whom
- ▸AI reasoning: each agent's analysis, citations to scientific papers
- ▸Disagreements: where agents disagreed and how consensus emerged
- ▸Supporting evidence: links to the studies backing each dimension
- ▸Alternative views: where credible disagreement exists in the literature
- ▸Update history: how scores change as new science emerges
Our Vision
Founded by Tom Kreynin, People for Food is built on a simple principle: trustworthy food health information should be a public good.
Inspired by Ethereum's thesis on public goods, we believe that critical information infrastructure shouldn't be controlled by corporations or governments. It should be owned by the people it serves.
We're partnering with the Metabolic Health Initiative to ensure that the latest science guides every decision.
The goal: make food health information transparent, decentralized, and honest. So that everyone—producers who want to compete on health, scientists who want to advance the field, and people who want to trust the food they eat—can participate in building a healthier food system.
The protocol is live. The system is transparent. The science is sound.