I just spent the last few weeks in the lab, and I have to share this breakthrough.
I was tired of hearing security teams complain about alert fatigue. I realized: if the attacks are autonomous, the defense must be, too.
That's why I focused on building AI Sentinel: Autonomous Cyber Defender. It's a fully serverless, multi-service architecture on Cloud Run—Frontend, API, and a separated AI Inference "Brain" all scaling independently.
The biggest win? I got it to perform autonomous triage and remediation on a threat, taking the response time from minutes to under 5 seconds—with zero human intervention.
The biggest lesson: Autonomy must equal Accountability. We built in mandatory guardrails and audit logs to ensure the AI's speed is matched by its safety.
Log in or sign up for Devpost to join the conversation.