NutriTail methodology

How the food suggestion works

NutriTail AI does not simply try to guess a food. It combines pet details, a structured food database, nutrition rules and a human explanation so the answer is useful and understandable.

In simple terms

How a food suggestion is produced

Customers do not need to see the technical machinery. They need to know that the suggestion is based on the animal, the food database and rules, not a random answer.

The pet comes first

The suggestion starts with species, weight, age, neuter status, activity, goal and what the pet likes or avoids.

The food database comes next

Foods come from Food V2 and the data we have collected, not from random AI memory.

Rules prevent mistakes

Allergies, the wrong species, age or size, and urgent concerns are filtered before reaching the customer.

AI writes clearly

OpenAI helps explain the result simply: what to choose, why, the starting grams and the next step.

Beta launch trust checklist

What you can rely on today and where we keep boundaries

NutriTail is designed for clear everyday guidance: a few practical food options, a starting amount and a next step. During beta, we clearly identify what is ready, what is improving and when a veterinarian is needed.

What is ready for beta

The login, chatbot, food suggestions, food selection, grams per day, saving, report and timeline flow is available for testing.

What remains under review

The food database, user feedback and edge cases continue to improve before a wider launch.

When a veterinarian is needed

For urinary or kidney concerns, diabetes, pancreatitis, blood, loss of appetite, severe vomiting or diarrhea, food does not replace veterinary guidance.

Beta user proof checklist

How a beta user helps improve real-world readiness

Progress improves when we prove that the flow works for real customers, not only because the code passes. These are the steps we want to see in every beta test.

  • Complete a full analysis from login to suggested foods.
  • Choose one food and check whether the grams per day make sense.
  • Save the analysis and open the pet report.
  • Return after 2-4 weeks for a progress check with weight, food grams, treats, stools and appetite.
  • Leave feedback if the suggestion was confusing, a food is missing or the pet did not accept the taste.

How we keep suggestions trustworthy

AI explains. NutriTail decides from data.

The suggestion is not based on random memory or generic assumptions. We first use pet details, then the Food V2 database, followed by nutrition and safety rules. OpenAI helps write the answer naturally without inventing foods or nutrient values.

1. Food V2 database

This is where the foods, calories, ingredients and nutrient data used in suggestions come from.

2. NutriTail rules

They remove unsuitable options based on species, age, size, allergies, weight goal and key health concerns.

3. OpenAI response

It turns the results into a clear human explanation without deciding which food is suggested on its own.

4. Safety boundaries

For urgent or medical cases, the flow stops the food-shopping path and recommends contacting a veterinarian.

AI boundaries

OpenAI helps with the conversation, not arbitrary food selection.

NutriTail controls the database, exclusion rules and final ranking. AI helps understand the customer's message and write the answer more naturally.

What is allowed

  • Explain clearly why a food may fit.
  • Ask only for missing information.
  • Turn rules and data into a human answer.

What is not allowed

  • It must not invent foods or nutrient values.
  • It must not ignore allergies, species or age.
  • It must not diagnose or make treatment promises.

Source quality

Not all data carries the same weight

The food database grows from official pages, PDFs, Greek e-shops and label photos. We keep the source of each data point so NutriTail knows when it can speak with confidence and when it should be more cautious.

1. Official source

We prefer an official company page, technical sheet or official PDF when available because these provide the clearest basis for nutrient and ingredient data.

2. Retailer or Greek e-shop

This is used to fill gaps or confirm products sold in the Greek market. A conflict with an official source requires review.

3. Label or package photo

This is highly useful for real shelf products, especially when online sources omit calories, ash, minerals or detailed ingredients.

4. Incomplete data

When critical details are missing, NutriTail does not pretend to be certain. The suggestion becomes more cautious and asks for more information when needed.

Commercial transparency

A brand must not be able to buy the suggestion.

NutriTail may offer plans, partnerships or featured options in the future. These must not bypass suitability rules: allergies, species, age, size, calories, weight goals and medical boundaries come first.

Rules come first

If a food does not match the species, age, allergy, size or a serious health goal, it must not be suggested because it is available or commercially important.

Partnerships do not hide the limits

Featured or promoted options may appear only within suitable results. They do not replace nutrition filters or remove warnings.

Customers need a clear reason

Every suggestion should explain clearly why it may fit: calories, fat, fiber, life stage, taste preference or sensitivity.

1. Understand the pet

It starts with species, age, weight, activity, neuter status, weight goal, sensitivities and taste preferences.

2. Calculate a practical target

It estimates daily calories, a treat allowance and a starting food amount you can put into practice.

3. Search the NutriTail database

Food suggestions come from our Food V2 database, not from random AI memory or generic answers.

4. Remove unsuitable options

It filters out species, life stage, size, allergy or nutritional signals that do not match the profile.

5. Explain the suggestion clearly

AI helps write the answer naturally, while NutriTail determines the foods and rules.

What the customer ultimately sees

The suggestion should be useful at home, not only correct in code.

NutriTail turns nutritional assessment into a simple flow: a few clear food options, the reason for each, a starting amount, saving and a follow-up check. Customers can remember what to do and return when weight, taste preference or results change.

1. A clear shortlist

The customer first sees a few strong options, usually premium and value choices, instead of a long technical list.

2. Why it was suggested

Each food includes a clear explanation based on age, weight, activity, neuter status, sensitivities, taste and available nutrient data.

3. Selection and grams per day

After the customer chooses a food, NutriTail calculates a starting amount in grams per day based on the calorie target.

4. Continue after the analysis

The analysis can be saved, turned into a report, added to a timeline and used later for a progress check or a new suggestion.

What we review in the food database

  • calories per 100 g
  • protein, fat, fiber and ash
  • calcium, phosphorus, magnesium and sodium when available
  • omega 3/6, EPA and DHA when listed
  • ingredients, health tags and nutrition use cases

Boundaries and safety

  • It does not replace a veterinarian or make diagnoses.
  • For urinary or kidney concerns, diabetes, pancreatitis, blood, severe vomiting or loss of appetite, it recommends a veterinarian.
  • When food data is missing, the answer becomes more cautious.
  • You can save a pet, view its history and run another progress check.

Ready for an analysis?

Start with your pet and save the analysis to your profile for a future progress check.

Start free