Reasoning LLM
A large language model project focused on Turkish natural language processing and medical reasoning; development continues on symptom analysis, reasoning and diagnosis suggestions.
Reasoning LLM is our R&D project aiming to build a large language model capable of medical reasoning in Turkish; work continues on step-by-step chain-of-thought reasoning, symptom analysis and diagnosis suggestions.
Key Features
Turkish NLP
Natural language processing stack specially trained for Turkish medical texts.
Medical Reasoning
Provides clinical assessment support by reasoning across findings.
Symptom Analysis
Analyzes patient symptoms and prioritizes possible diagnoses.
Diagnosis Suggestions
Offers decision-support diagnosis suggestions; the final call always belongs to the physician.
Step-by-Step Reasoning
Builds chain-of-thought reasoning such as "headache + fever + neck stiffness → differential diagnosis → urgency".
Flexible Training Strategy
Training from scratch on Turkish medical text, or specializing open-source models with LoRA/QLoRA — both paths are being evaluated.
Large-Scale Target
A reasoning-capable model with 70B+ parameters is targeted for medical decision support; training is planned on NVIDIA DGX-class infrastructure.
Integration with Health Products
Connects to WebAHBS, the smart kiosk and pre-assessment flows through a central AI-API; suggestions are applied with physician approval.