AEGIS is a self-evolving WAF
Static WAFs only block attacks someone already wrote a signature for. AEGIS closes that gap autonomously. Traffic our rules can't explain gets embedded and clustered on an AMD Instinct MI300X, then sent to an LLM (also GPU-hosted) that proposes a regex and rationale. A deterministic Rust engine tests the candidate against real attack and benign corpora nothing ships on an LLM's opinion alone. Validated rules hot-reload live, no restart, with a human-review mode available. Underneath sits a production-grade Rust/Pingora core: OWASP-aligned rules, DDoS/rate limiting, sub-30ms decisions, and a dashboard with live gauge telemetry per AI signal. Three AMD GPU workloads power it, LLM inference, continuous model training, large-batch embedding benchmarked against CPU, not just claimed.
Category tags:
.png&w=96&q=75){kind=small}
For Innovators & Creators