Anandhasasidharan S
Cybersecurity researcher · AI/ML engineer · Builder
B.Tech CSBS @ SRM Institute of Science and Technology
locate --interests
robotics × artificial_intelligence × hacking × interpretability
ps aux | grep active_projects
- Network Anomaly Risk Score (ML threat detection + XAI)
- Community AI Audit Framework (multi-provider LLM auditing)
- Pretext DOM-free text layout engine
- Autonomous agent infrastructure research
$ _
The Mission
I believe that security AI must be interpretable. Black-box models making security-critical decisions are a liability, not an asset. Every alert must be traceable from model output to raw evidence.
I'm building Small Language Models (SLMs) and agentic frameworks specifically for blue-team cybersecurity operations. The goal: autonomous agents that can explain every decision they make — full reasoning traces, structured audit logs, and post-hoc interpretability.
Stack & Tools
Connect
I'm interested in security research, interpretability work, and practical systems that need both engineering and taste.
Schedule
Want to talk security research, interpretability, or a potential collaboration? Check my availability below and grab a slot.