How FAMMO Works

FAMMO is an AI-powered nutrition engine that generates personalized meal plans and health guidance for dogs and cats. It processes structured pet data, applies veterinary nutrition science, and produces deterministic JSON outputs for meal planning and health reporting.

1. Input Layer — Pet & Owner Data

The system starts from structured pet profiles collected through our web interface. Each profile captures comprehensive data points that form the foundation for personalized nutrition planning.

Pet Profile Inputs

• Species (dog/cat)
• Breed
• Age & life stage
• Weight & body condition
• Activity level
• Allergies
• Health conditions
• Feeding history

Owner Preferences

• Feeding goals (weight loss, maintenance, growth)
• Treat usage preferences
• Veterinarian recommendations
• Food type preferences

2. Feature Engineering — Turning Profiles into Nutrition Targets

Raw pet data is transformed into structured nutritional features through our feature engineering pipeline. This layer applies veterinary formulas and evidence-based logic to calculate precise dietary requirements.

Caloric Needs

MER/DER calculation based on weight, age, and activity

Nutrient Ratio Targets

Protein, fat, and carbohydrate percentages

Ingredient Safety Mapping

Allergen exclusions and toxic food identification

Breed-Specific Adjustments

Tailored recommendations for breed characteristics

Health Flags

Overweight, senior, sensitive digestion markers

Portion Scaling

Precise quantities based on weight and DER

3. AI Decision Layer — How Meals Are Generated

The AI engine combines veterinary nutrition rules with machine-learning logic to generate safe, balanced meal plans. This layer ensures compliance with nutritional guidelines while respecting all constraints.

1

Calculate Daily Caloric Requirement

Apply MER/DER formulas based on pet profile
2

Apply Constraints

Filter for allergies, safety rules, breed considerations, and age requirements
3

Generate Meal Plan Options

Create 1–2 diverse meal plans with balanced variety
4

Build Daily Structure

Distribute calories across breakfast, lunch, dinner, and treats
5

Validate Nutrient Balance

Ensure protein, fat, and carbs meet target percentages
6

Attach Safety Notes

Include feeding tips, transition guidance, and monitoring alerts

4. Structured JSON Output

All outputs are deterministic JSON structures, ensuring reliability, reproducibility, and seamless integration with partner systems. This structured format enables API consumption by veterinary clinics, pet food companies, and third-party applications.

{
  "der_kcal": 240,
  "nutrient_targets": {
    "protein_percent": "35%",
    "fat_percent": "20%",
    "carbs_percent": "45%"
  },
  "options": [
    {
      "name": "Option 1",
      "overview": "Balanced diet using both dry and wet food.",
      "sections": [
        {
          "title": "Breakfast",
          "items": ["30g high-quality kitten dry food"]
        },
        {
          "title": "Lunch",
          "items": ["40g premium kitten wet food (chicken)"]
        },
        {
          "title": "Dinner",
          "items": ["30g high-quality kitten dry food"]
        }
      ]
    }
  ],
  "feeding_schedule": [
    {
      "time": "8:00 AM",
      "note": "Serve breakfast with fresh water."
    }
  ],
  "safety_notes": [
    "Ensure fresh water is available at all times.",
    "Introduce any new foods gradually."
  ]
}

API Ready: This JSON structure can be consumed via REST API endpoints, enabling clinics and partners to integrate FAMMO's nutrition intelligence into their own systems.

5. Constraints & Safety Rules

Hard Constraints

✓ Calorie range must match DER ±5%
✓ All allergens must be excluded
✓ Species-specific safety rules enforced
✓ No toxic ingredients (onions, chocolate, grapes, etc.)

Soft Constraints

◆ Palatability and taste preferences
◆ Variety and ingredient rotation
◆ Owner budget and availability preferences
◆ Food type preferences (wet/dry/raw)

Clinical Grade: This multi-layer constraint system makes FAMMO suitable for real clinical use, ensuring that every meal plan meets both safety requirements and practical feeding considerations.

6. System Architecture (High-Level)

FAMMO is built on a robust, scalable technology stack designed for reliability and integration flexibility:

Django Backend

Python web framework with ORM and REST API

AI Engine

Python-based nutrition logic and ML models

PostgreSQL Database

Relational data storage for pets, users, and plans

Tailwind UI

Modern, responsive user interface

JSON Outputs

Structured data for consumption

REST API

Partner integration endpoints

Data Flow Pipeline

Pet Profile

Input Layer

Feature Engineering

Transformation

AI Decision Layer

Generation

JSON Output

Structured Data

UI & API

Delivery

7. Why FAMMO is Deep-Tech

Deterministic AI Outputs

Structured JSON ensures reproducibility and reliability

Evidence-Based Veterinary Logic

Grounded in clinical nutrition guidelines and research

Multi-Layer Constraints

Complex feature engineering and safety validation

Clinic & Partner Integration

REST API ready for B2B partnerships and white-label use

Built to Scale Across EU

Multi-language support and localized nutrition guidelines

Continuous Learning System

Feedback loops improve accuracy and personalization over time

Investment Opportunity: FAMMO combines proprietary AI technology with clinical-grade nutrition science, creating a defensible moat in the pet health market. Our structured approach enables rapid scaling and B2B partnerships across the European Union.