Ernest Tretjakov
Built at Meta Global AI Developer Hackathon · Oct 26, 2026 · Remote

JA4+ Network Forensics — Turning Network Traffic into Explainable Cyber Threat Intelligence As cyberattacks grow increasingly sophisticated and security vulnerabilities continue to emerge, organizations face a critical challenge: detecting suspicious network activity is not enough — security analysts must understand what happened, why it matters, and how to respond. Traditional security tools often generate overwhelming amounts of alerts, making investigations time-consuming and requiring specialized expertise. JA4+ Network Forensics addresses this problem by combining network traffic analysis, behavioral anomaly detection, and fine-tuned Meta Llama 3.2 AI into a unified investigation platform. Instead of simply identifying anomalies, our system connects suspicious communication patterns, investigates potential threats, and generates evidence-based explanations with actionable recommendations. Our innovation transforms complex network data into understandable security intelligence, helping analysts investigate threats faster, reduce manual analysis, and make advanced AI-powered cybersecurity more accessible, affordable, and privacy-conscious.