Inzex Forensics

Inzex Forensics is an autonomous Digital Forensics and Incident Response (DFIR) platform built for solo cybersecurity consultants and Managed Service Providers who cannot afford enterprise tools like CrowdStrike. Today, analyzing a memory dump requires an analyst to manually run 8-12 Volatility 3 plugins from the command line, cross-reference thousands of lines of terminal output, and spend 4-6 hours writing a report. Inzex compresses this into 3 minutes. The analyst uploads a raw memory dump through a clean dark-mode web interface. The backend, running on AMD Developer Cloud with an AMD MI300X GPU, automatically executes Volatility 3 plugins including pslist, malfind, netscan, and cmdline. The structured output is piped directly into Gemma 4 12B-it, running locally on AMD ROCm hardware via vLLM 0.23, where it acts as a Senior SOC Analyst — identifying anomalies, mapping each finding to a MITRE ATT&CK technique, assigning confidence scores, and citing the exact Volatility output line behind every claim. The findings populate a Human Review Workspace where the analyst can confirm, edit, or reject each AI inference individually. Nothing enters the final report without passing through human review. The analyst can also submit feedback on any finding, triggering a live re-evaluation by Gemma 4 that updates the UI in real time. One click exports a professional, dark-branded PDF incident report ready to hand to a client. Raw evidence never leaves the AMD processing environment. Files are processed in an isolated temp directory and permanently deleted after analysis. Only the structured findings report persists in Supabase. Built on: AMD MI300X GPU, ROCm 7.2, vLLM 0.23, Gemma 4 12B, Volatility 3, Next.js, Supabase, Netlify.