Inspiration
Cybersecurity threat auditing is often complex, heavy, and limited to resource-intensive environments. We wanted to build a lightweight, intelligent host-based security agent that can run instantly in any cloud-native environment with zero local configuration friction.
What it does
SentinelAI Core ingests simulated system audit telemetry, runs high-frequency behavioral heuristics, identifies anomalous privilege escalations or unauthorized modifications, and exports structured JSON reports for security operations.
How we built it
Built entirely in Python, leveraging native structured data processing and algorithmic risk-scoring pipelines. Designed for seamless execution inside Google Colab to guarantee accessibility and zero-dependency friction.
Challenges we ran into
Balancing the lightweight execution constraints of a cloud notebook with realistic threat-detection scoring heuristics, ensuring clean JSON telemetry outputs without external bloat.
What we learned
We mastered rapid prototyping for security auditing agents and optimized automated log analysis workflows tailored for rapid hackathon delivery.
