Built an explainable AI credit decision platform using LangGraph, Qdrant, and RAG architectures. The system uses multiple AI agents to analyze applications, retrieve relevant information, assess risks, and generate transparent decisions. The platform combines vector search, agent orchestration, and explainable AI to improve auditability and trust. Key features: • Multi-agent workflows • Explainable AI decisions • Hybrid vector search • Risk assessment • Audit-ready outputs