ZKIP-CLINICAL — Clinical Decision Support Platform

A clinical decision support platform built for Kottackal Business Solutions, combining on-chain patient credentialing, a graph database of clinical relationships, and an AI risk-retrieval layer.

• Combined on-chain patient credentialing with a Neo4j graph of drug-disease-allergen relationships and an AI risk-retrieval layer seeded from RxNorm, MED-RT, PharmGKB, SNOMED CT, and ICD-10-CM across 300,000+ records.
• Engineered a custom vector-based clinical memory service that retrieves patient risk profiles using semantic similarity search across pharmacological, therapeutic, and demographic signals.
• Integrated Corti AI for clinical NLP — automated PDF-to-diagnosis coding, real-time drug contraindication detection, and guardrail enforcement.
• Implemented zero-knowledge proof circuits for privacy-preserving identity and prescription verification, and ERC-721 smart contracts for tamper-proof on-chain credentialing.
• Demonstrated interpretable AI behaviour: removing a single patient allergy record causally shifted the system's risk classification, proving traceable, auditable AI decisions.