AI Security Engineer · Cybersecurity
Specialist · Cloud Architect.
⚙️ Mechatronics Engineer (81% First Class) → B.E.
CSE Cybersecurity.
Final-year B.E. CSE (Cybersecurity) at Saveetha Engineering College
2026.
IBM Z Datathon 2025 Runner-Up · 2 IEEE
Publications · 50+ Certs.
I'm a final-year B.E. Computer Science (Cybersecurity) student at Saveetha Engineering College, Chennai — graduating 2026. My unique background starts with a Diploma in Mechatronics Engineering (81% First Class with Distinction) at TVS CPAT, giving me rare cross-domain hardware + software thinking that most engineers never develop.
From the IBM Z Datathon 2025 national 2nd place (Heart MRI AI on IBM Mainframe) to IEEE-published research on AI-powered security systems, I build production-grade work that solves real problems — not just academic demos.
Currently deepening expertise in LLM fine-tuning (Gemma3 + JAX), cloud-native security architectures, and adaptive threat detection. My mission: build AI systems that don't just defend — they adapt and outthink attackers.
AI-Powered Heart MRI Classification on IBM Mainframe. Team BIG IRON. <2s inference. Grad-CAM XAI explainability.
$500 + IBM Mentorship + Lab VisitWAF AI — Conference + Journal (Published). CSSS — Accepted with 100/100/100 review scores. AI Security + Medical Imaging.
Peer Reviewed · AcceptedMechatronics Diploma (PLC, CNC, Robotics, AutoCAD) + B.E. Cybersecurity. Industrial training at Brakes India, MK Auto, RETECH Solutions.
The Combo Nobody Else HasNational award-winning CNN on IBM Z Mainframe. Grad-CAM + Integrated Gradients XAI explainability. IBM Cloud Object Storage & Db2. <2s inference. $500 prize + IBM mentorship + Bangalore Labs visit.
ML WAF blocking SQLi, XSS, CSRF & zero-days in real-time. Dual ML engine (Isolation Forest + Random Forest). Auto nginx rule generation. Prometheus/Grafana/Loki observability. 8+ Docker containers.
89.51% accuracy on 217,875 medical images. MobileNetV2 + 4-role clinical workflow (Patient/Doctor/Pharmacist/Admin). Auto PDF reports. JWT + OTP 2FA + RBAC. Reviews: 100/100/100.
U-Net architecture for automated cardiac MRI segmentation (LV, RV, Myocardium). 94.8% accuracy. 6-class disease classification. Reduced diagnosis from hours to <60 seconds.
Fine-tuned Gemma3 1B using JAX-native Tunix library for step-by-step LLM reasoning. Full RLHF/RLAIF pipeline on TPU. 322 teams · 11,173 entrants · $100K prize pool.
Cross-platform Electron app. CPU/GPU/RAM/SMART detection, health score 0–100, driver audit, performance tier. Built for real users who diagnose their own machines.
Obstacle detection 2cm–400cm (±0.3cm). Silent haptic feedback with 3-zone vibration. Piezoelectric self-charging from walking. 17-day battery life. Led 6-member team.
ATS score, keyword gap analysis, resume builder (4 templates), LinkedIn scraping. Built collaboratively with Vishwa Rathinam using Google Gemini API and spaCy NLP.
9 Courses · Google / Coursera
Avg 96% · Sep 4, 2025Professional Certificate · 4+1 Courses
All 99–100% · Mar 30, 2025Specialization · LearnKartS · 4 Courses
Avg 94% · Aug 25, 2025Specialization · LearnQuest · 4 Courses
Avg 92% · Aug 22, 2025IBM Z Platform Badge
Nov 12, 2025Virtual Experience · Forage
Sep 1, 2024Press ` for the command palette
Actively seeking full-time roles in AI Security, Cloud Security, and Cybersecurity. Open to research collaborations, internships, and innovative projects.