Suricata, Supercharged.

A powerful bolt-on toolkit that transforms Suricata into a capable NDR with AI-powered detection and automated response.

Don't replace your stack. Extend it. Rockfish bolts onto your existing Suricata deployment. No rip-and-replace. No vendor lock-in. No SaaS dependency. No hype. Just tools that works.

Bolt-on

  • Drops into your existing Suricata deployment

  • Integrates with your SIEM, SOAR, and ticketing systems through n8n

  • MQTT event bus connects to anything in your stack

  • Deploys on any commodity hardware — no proprietary appliances

  • Air-gap ready: fully disconnected operation, no cloud required

  • No recurring subscription. No SaaS to kiss.

Detect

  • Suricata integration with full EVE JSON enrichment

  • Unsupervised ML for signature-free anomaly detection

  • Encrypted traffic analysis via JA3/JA4 fingerprinting

  • AI-ready: query network data conversationally via MCP server

Respond

  • AI-assisted Suricata rule injection on detection

  • Workflow orchestration with n8n for custom response playbooks

  • SIEM/SOAR forwarding with fluentbit to Splunk, Sentinel, Shuffle

  • MQTT event bus for real-time alert distribution

Archive

  • Full metadata capture including enriched flow.

  • Arrow Parquet for fast analytical queries at scale.

  • Store local or push to S3-compatible storage.

  • Meet regulatory and audit requirements with immutable storage.

  • Analyze historical data with any Parquet compatible engine.

How It Works

  • Capture — Suricata passively monitors your network via TAP/SPAN, feeding traffic to Rockfish.

  • Detect — Rockfish's behavioral engine analyzes enriched metadata to identify lateral movement, C2 beacons, data exfiltration, and anomalies that signatures miss.

  • Respond — Detection events fire over MQTT to n8n, triggering automated playbooks.

  • Archive — Every flow is enriched and stored in immutable Parquet format.


Your network. Your insight. Your archive.

Coming Q2 2026. Preview the docs, explore the architecture, or get in touch.


© 2025-2026. Fidelis Machines, LLC. All rights reserved.