❓ Frequently Asked Questions
What is MedMemory AI?
MedMemory AI is an AI-powered personal medical memory system that transforms medical documents into a searchable health timeline and conversational medical history.
Is MedMemory an Electronic Health Record (EHR)?
No.
MedMemory focuses on building an AI-powered medical memory layer rather than replacing traditional EHR systems.
What documents can be uploaded?
Currently:
- Prescriptions
- Lab Reports
- Scan Reports
- Discharge Summaries
Does MedMemory store uploaded PDFs?
The platform primarily stores extracted structured medical information and semantic memory representations.
Storage strategies may evolve as the project matures.
Which AI models are used?
By default:
qwen2.5:3b
mxbai-embed-large
through Ollama.
Which databases are used?
PostgreSQL
Stores:
- Patient records
- Medical documents
- Timeline information
Qdrant
Stores:
- Embeddings
- Semantic memory
Can I self-host MedMemory?
Yes.
The project is designed with self-hosting in mind.
Deployment guides are available in the documentation.
Does MedMemory require internet access?
For local deployments using Ollama:
- Document processing can run locally.
- Chat functionality can run locally.
Cloud deployments may require internet access for managed services.
Is MedMemory production ready?
Currently the project is under active development and should be considered experimental.
How can I contribute?
See:
Guides → Contributing Guide
for contribution instructions.
Where can I report bugs?
Open an issue on GitHub:
https://github.com/MedMemory-AI/MedMemory/issues