How dastakfatemeh/pet_vetclinic works
The product is a FastAPI pet-health assistant: an owner submits symptoms, the API classifies them, retrieves similar cases when confidence is sufficient, and generates plain-language findings. A separate response endpoint handles appointment intent and provides a scheduling link. The graph also shows Qdrant storage and the VetBERT embedding layer. Relationships reflect the excerpts and README; unsampled implementation details are omitted.
Main entry points: AI_Agents.py FastAPI application · RetrievalAgent.py Similar-case retrieval · ClassificationAgent.py Condition classification · VetBertMixin.py VetBERT embeddings
Components 10
API and orchestration
- FastAPI application
AI_Agents.py - Appointment response
AI_Agents.py
Clinical analysis
- Condition classification
Agents/ClassificationAgent.py - Similar-case retrieval
Agents/RetrievalAgent.py - VetBERT embeddings
Agents/VetBertMixin.py
Communication
- Clinical explanations
Agents/CommunicationAgent.py
External services
- Qdrant vector store
- Hugging Face models
- Scheduling service
Other components
- Pet owner
Connections 9
- Pet owner → FastAPI application: submits symptoms
README.md - FastAPI application → Condition classification: classifies symptoms
AI_Agents.py - FastAPI application → Similar-case retrieval: requests cases
AI_Agents.py - Similar-case retrieval → VetBERT embeddings: generates vectors
Agents/RetrievalAgent.py - Similar-case retrieval → Qdrant vector store: searches cases
Agents/RetrievalAgent.py - FastAPI application → Clinical explanations: explains findings
AI_Agents.py - FastAPI application → Hugging Face models: loads models
AI_Agents.py - Pet owner → Appointment response: answers prompt
README.md - Appointment response → Scheduling service: provides booking link
AI_Agents.py