VetApp Framework in numbers
4
universities
research universities participating in the creation of the framework
0
hallucinations
deterministic inference
7
layers
VetApp technological layers
4
products
products on one platform
Built and verified with leading institutions
The VET-MED Framework is a scientific platform developed with four research universities.
Verification of medical knowledge base, veterinary ontologies
Clinical validation, patient case datasets
Neuro-symbolic architecture, inference methods
Language Models, Entity Extraction, and Normalization
KNOWLEDGE LAYER
How is the VET-MED AI Framework structured?
We didn't build products based on other people's LLMs via API. We built our own specialized AI platform, and our products are a natural application of it.
Medical Knowledge Datasets
Veterinary knowledge base – diseases, symptoms, diagnoses, and treatments. In collaboration with Cornell University and ETH Zürich.
Nutrition Knowledge Datasets
Nutrients, standards per species and breed, interactions with drugs and diseases.
Patients Cases Datasets
30k+ triage analyses and 25k+ chat consultations. Every new user strengthens the model.
PROCESSING LAYER
Medical Entity Extraction
Recognizes medical entities from owner description: symptoms, medications, procedures, anatomy.
Veterinary Entities Normalization
Normalizes entities to standard veterinary ontologies.
Neuro-Symbolic Reasoning Model
The heart of the framework. It combines neural networks with symbolic rules. Every conclusion is traceable to the knowledge base. Zero hallucinations.
Neural Small Language Model
A language model trained on veterinary data that supports user communication.
PRODUCTS
Why does this matter for regulated environments?
In insurance and veterinary medicine, "the model was wrong" is not an acceptable answer.
Hallucinations in diagnosis
A generative model might conclude that a nonexistent drug cures a nonexistent disease. In medicine, this is disqualifying.
Black box
You can't explain why the model made a particular decision. In insurance, every decision must be auditable.
Inconsistent responses
The same question yields different answers in different sessions. This is a disaster for the claims process.
Deterministic inference
The neuro-symbolic model infers from a knowledge base, not generates it. Each response is traceable to clinical facts.
Full interpretability
You can explain every decision step by step. This is crucial for insurance audits.
Always the same answer
The same inputs produce the same output. Essential for reliable decision-making.
Download the VetApp application
With VetApp, taking care of your pets has become even easier and more convenient. See for yourself!
Download the VetApp application.
Choose a subscription plan that suits you and your pet's needs.
Take your pet care to the next level.
Click on your store icon and go to the app.
.png)
.png)