Built for Intelligence
Most software is built to store text. V.E.T.S. is built to store meaning.
Our architecture is designed from the ground up to capture the complex, interconnected nature of animal biology and management. It is robust, scalable, and ready for the AI era — because AI wasn’t bolted on, it was planned for.
The Semantic Foundation
Traditional databases store flat records. V.E.T.S. stores knowledge in a structure that mirrors how experts actually think:
Traditional Database
patients (id, name, owner_id)
visits (id, patient_id, date, notes)
treatments (id, visit_id, drug, dose)
-- Flat tables. No context. No connections.
Find a patient. Read their visits. Read their treatments. The relationships are mechanical. The meaning is lost.
V.E.T.S. Architecture
Items (unified: animals, docs, procedures)
TreeValues (hierarchical relationships)
DynamicForms (context-aware evaluation)
-- Everything connects. Context preserved.
Every animal is an Item. Every treatment is an Item. The tree structure captures WHY they’re related, not just THAT they’re related.
AI-Ready Architecture
Five design decisions made years ago now enable V.E.T.S. to integrate AI more deeply than any competitor:
1
Everything is an Item
The unified stbl_System_Items table means the AI can traverse from an animal to its procedures to related literature through a single relationship system. No joins across incompatible schemas.
2
HTML-in-SQL Security
Stored procedures generate HTML directly, meaning AI-powered features inherit all 6 security layers automatically. The AI can’t see data the user doesn’t have permission to see.
3
TreeValue Hierarchies
The tree structure gives AI context that flat databases lack. An AI query about “lameness in quarter horses” can traverse the breed tree to find relevant protocols, related conditions, and cross-referenced cases.
4
TeamDoc Content Management
Knowledge lives in TeamDoc — a structured yet flexible system that predates RAG but contains everything RAG needs: hierarchical documents, chunked content, relationship tables, and human-friendly editing. It was RAG before RAG had a name.
5
Multi-Tenant by Design
Row-level security means AI features work across clients without any data leakage. A vet clinic’s AI minions see only that clinic’s data, even though all clients share the same database infrastructure.
Why This Matters
These architectural decisions have compounding benefits:
For Veterinarians
AI assistants that understand your practice’s protocols, your patients’ histories, and your preferred treatment approaches — not generic responses from a model trained on the internet.
For Developers
A consistent, well-documented architecture where AI capabilities are added through stored procedures, not external services. Database is the single source of truth.
For the Industry
A platform that can grow to serve the entire animal care ecosystem while maintaining strict data isolation, regulatory compliance, and audit trails.
Architecture as Security: V.E.T.S. doesn’t bolt security on after the fact. The HTML-in-SQL pattern means data never leaves SQL Server without passing through permission checks. The Items table unification means AI can’t access a document it shouldn’t see — because the same security layers that protect a patient record also protect every knowledge chunk the AI retrieves.
Built to Last, Built to Learn
The foundation has been production-stable since 2009, serving real clinics with real animals. The AI layer is being built on that proven foundation — not replacing it, but amplifying it.