I want to build a knowledge-based decision support system specifically for clinical use. The platform must accept core patient data—symptoms, vitals, basic labs—then use an explicit rule set to deliver three things: probable diagnoses ranked by confidence, evidence-based treatment recommendations, and concise documentation snippets suitable for pasting straight into an EMR. A classical expert-system architecture makes the most sense here. I will supply—or work with you to source—standard medical ontologies such as SNOMED-CT, ICD-10, and up-to-date drug interaction tables. Your task is to encode them into a transparent rule base and wire that to an inference engine (Drools, CLIPS, Prolog, or a modern Python rules framework are all acceptable) exposed through a lightweight web UI or a RESTful API. Every output must be traceable; the system should show which rules fired and why. Deliverables • Complete, well-commented source code and knowledge base • Docker-ready deployment script or equivalent one-step installer • Sample dataset plus unit tests demonstrating correct diagnosis and treatment mapping • Quick-start user guide Acceptance criteria: installation on an Ubuntu VM without manual patching, ≥85 % accuracy on the provided test set, and clear rule-trace explanations for each suggestion. Include a concise milestone plan and any preferred tooling in your proposal. ** I have full product description prompts