TRiaD
TRiaD (Threat Intelligence & Automated Defense) is an AI-powered cybersecurity incident response platform developed during the Band of Agents Hackathon 2026. The system leverages a collaborative multi-agent architecture to automate the traditionally time-consuming processes of threat analysis, incident triage, intelligence correlation, and compliance reporting. The platform consists of specialized AI agents working together in a coordinated workflow. Incoming security alerts are processed by an ingestion layer, enriched with contextual threat intelligence, and analyzed through semantic similarity searches using a vector database. The analyst agent investigates indicators of compromise, correlates findings with historical incidents and MITRE ATT&CK techniques, and generates actionable insights. A manager agent then compiles compliance-ready incident reports suitable for security operations centers and organizational stakeholders. TRiaD provides a modern web dashboard with real-time updates, interactive alert monitoring, and automated reporting capabilities. By combining FastAPI, Next.js, ChromaDB, Gemini-powered reasoning, and WebSocket-based communication, the platform demonstrates how autonomous AI agents can significantly accelerate cyber defense operations while maintaining transparency, traceability, and auditability. The project showcases practical applications of agentic AI in cybersecurity, threat intelligence automation, incident response orchestration, and security analytics.
Category tags:.png&w=96&q=75){kind=small}
.png&w=96&q=75){kind=small}
"Application of Technology: 3/5 — TRiaD uses Band as the coordination mechanism for its three-agent triadic debate structure, but the Band integration is functional rather than advanced. The three-agent adversarial setup is well-conceived, though Band's more powerful routing and control features could be better leveraged. Presentation: 4/5 — Good demo video showcasing the triadic debate process with clear agent roles. The README explains the concept well and the visual presentation of agents debating is easy to follow. Business Value: 4/5 — Triadic debate and adversarial agent reasoning has strong applicability in regulated decision-making contexts (legal, medical, financial) where multiple perspectives and challenge/counter-challenge are required. Originality: 3/5 — The triadic debate concept is intellectually interesting, though adversarial multi-agent systems are an established pattern. The specific three-agent structure is a creative constraint that adds some novelty. Overall: TRiaD implements a structured adversarial debate protocol between three agents coordinated via Band. The concept is sound and has real applications in high-stakes decision environments. The execution is solid with a working demo, but the Band integration could go deeper and the business use cases could be more concretely demonstrated."
Sanem Avcil
For Innovators & Creators