cyber-doctor

Report Abuse

Basic Information

cyber-doctor is an open-source multi‚Äëmodal health assistant project that integrates large language models, retrieval augmentation and knowledge graph techniques to provide basic medical triage, medical record analysis and professional Q&A. The repository bundles a Gradio web interface, multi‚Äëmodel orchestration and connectors for several model providers so users can run a local or API‚Äëbacked assistant that accepts text, images, audio and uploaded files. It targets users who care about personal health and developers who want to adapt a medical domain agent to other specialties by swapping knowledge bases or models. The project emphasizes deployment options, configurable APIs, RAG workflows and optional Neo4j knowledge graph support to improve domain accuracy.

Links

Categorization

App Details

Features
The codebase implements multi‚Äëmodal inputs (text, image, audio, file upload) and multi‚Äëmodel decision logic that routes tasks to appropriate models. It includes voice dialogue with STT and TTS (Whisper and edge‚Äëtts), image and video generation capabilities, automatic PPT/Word document generation, multi‚Äëturn conversation memory, RAG retrieval from local knowledge bases and internet search, and optional Neo4j knowledge graph integration. The repo uses LangChain, PyTorch, Transformers, ModelScope/Hugging Face support and a Gradio front end. Modular packages cover clients, RAG services, knowledge graph matching, document generation and internet scraping, enabling customization and API adapter extensions.
Use Cases
cyber-doctor helps widen access to basic medical guidance by enabling users to get preliminary disease suggestions, extract and analyze information from medical records and drug labels, and ask specialty questions with retrieval‚Äëenhanced answers. It automates document and presentation generation for clinical summaries, supports voice interactions to lower entry barriers, and can combine up‚Äëto‚Äëdate internet retrieval with curated knowledge graphs to improve answer relevance. For developers and educators it offers a working example of integrating RAG, multimodal inputs and Neo4j to build domain experts; for underserved regions it offers a local deployable assistant to reduce dependence on remote specialist visits.

Please fill the required fields*