Guardian Nexus

Guardian Nexus is an AI-powered digital guardian designed to help users identify and understand scams and digital threats in real time. The system starts with live audio, which is converted into text using the AssemblyAI Realtime API. The transcript is then processed through a LangGraph-based multi-agent workflow. Specialized agents analyze the conversation for scam patterns, suspicious requests, impersonation, manipulation, and other potential security threats. Instead of relying entirely on an LLM to make the final decision, Guardian Nexus combines AI reasoning with a deterministic 0–100 risk engine. This provides a consistent risk assessment while the LLM acts as an explanation and reasoning layer, converting technical analysis and security signals into clear, human-readable explanations. The backend is built with Python and FastAPI, with WebSockets enabling real-time communication between the backend and frontend. The frontend uses Next.js, React, TypeScript, and Tailwind CSS to provide a real-time security dashboard. The system also includes provider abstractions for AssemblyAI, structured events, asynchronous processing, persistence, error handling, and extensive automated testing. More than 398 tests have been passed across backend services, agent workflows, risk-engine logic, AssemblyAI providers, real-time functionality, integrations, and error handling. The long-term vision is to expand Guardian Nexus into a broader digital safety platform with mobile protection, trusted-contact alerts, family protection, stronger deepfake detection, persistent threat memory, and cross-device security.