VitalPet-RD

AI assistant for companion animal nutrition and pet-product R&D.

VitalPet-RD is a domain-specialized AI assistant designed to support research and development of pet nutraceuticals, functional foods, supplements, and related products for dogs and cats.

It is built on Qwen3-4B and optimized through domain-specific system instructions and an evidence-first R&D framework.

Quick Start

Install Ollama:

https://ollama.com

Then:

ollama pull katiealex/vitalpet-rd:0.1
ollama run katiealex/vitalpet-rd:0.1

What VitalPet-RD Can Help With

Ingredient Research

• Functional ingredient analysis
• Mechanisms of action
• Dog vs. cat differences
• Dose-range reasoning
• Ingredient compatibility
• Safety considerations

Formulation Development

• Nutraceutical formulations
• Functional food concepts
• Dosage-form development
• Palatability considerations
• Stability considerations
• Manufacturing feasibility
• Cost-oriented formulation analysis

Scientific Evidence

VitalPet-RD is designed to distinguish between:

• Established evidence
• Limited evidence
• Indirect evidence
• Hypotheses
• Areas requiring additional verification

It should not fabricate scientific papers, citations, doses, regulatory requirements, or clinical evidence.

Veterinary and Pet Health Topics

The model can assist with research-oriented questions involving areas such as:

• Feline and canine kidney health
• Joint health
• Dental health
• Digestive health
• Stress and behavioral support
• Immune and nutritional support

The model is intended for R&D support, not diagnosis or treatment.

Evidence-First Approach

VitalPet-RD follows an R&D workflow:

Evidence
   ↓
Mechanism
   ↓
Dose rationale
   ↓
Formulation
   ↓
Safety
   ↓
Palatability
   ↓
Stability
   ↓
Manufacturing
   ↓
Cost
   ↓
Regulatory considerations
   ↓
Claims

A key design principle is:

When reliable canine or feline evidence is insufficient, the model should state the limitation rather than generate a precise-looking but unsupported answer.

Dog vs. Cat

VitalPet-RD attempts to distinguish species-specific considerations rather than automatically transferring human or canine evidence to cats.

Species, life stage, dose, formulation, and route of administration should be considered when evaluating evidence.

Example Questions

What is the evidence for UC-II in dogs and cats?

Compare STAY-C 50, zinc, lysozyme, and bromelain for a feline dental formulation.

How should a low-phosphorus renal supplement be formulated for cats?

Evaluate this joint-support formulation for dogs and identify potential formulation issues.

What evidence exists for an ingredient in feline CKD?

Which claims for this pet supplement require additional evidence?

Limitations

VitalPet-RD is an AI research assistant. It may produce incorrect or incomplete information.

Scientific, veterinary, regulatory, dosage, safety, and clinical decisions should be independently verified against authoritative sources and qualified professionals.

The model does not replace veterinary diagnosis or professional scientific/regulatory review.

Model Information

• Model: VitalPet-RD
• Version: 0.1
• Base model: Qwen3-4B
• Purpose: Companion animal nutrition and pet-product R&D
• License: See the project repository and applicable upstream model license.

Project

GitHub:

https://github.com/katiealex/vitalpet-rd

The GitHub repository contains project documentation, evaluation materials, and development information.

Roadmap

Future versions may include:

• Ingredient knowledge database
• Scientific literature retrieval
• Evidence verification
• Formula calculation tools
• Ingredient compatibility analysis
• Regulatory/claims analysis
• Private R&D data integration
• Automated formulation evaluation

Disclaimer

VitalPet-RD is provided for research and development purposes.

It does not provide veterinary diagnosis, medical treatment, or regulatory/legal advice. Users are responsible for independently verifying scientific, safety, veterinary, manufacturing, and regulatory information before making product or clinical decisions.