Aarav Shah

Cybersecurity & AI/ML Specialist

Chennai, IN.

About

Highly accomplished Bachelor of Technology student specializing in Computer Science, with a 9.07 CGPA and practical experience in cybersecurity and AI/ML. Proven ability to develop robust threat detection pipelines, analyze vast datasets for vulnerability intelligence, and engineer advanced AI models, as demonstrated through internships at Mphasis and CyberXtron. Eager to leverage strong technical skills and research capabilities in impactful cybersecurity or AI/ML engineering roles.

Work

Mphasis

|

Threat Hunting and ML Intern

Bangalore, Karnataka, India

→

Summary

As a Threat Hunting and ML Intern at Mphasis, Aarav configured security tools, developed detection rules, and streamlined incident response workflows.

Highlights

Configured and maintained SIEM platforms (Splunk/Wazuh), AV, and monitoring tools, developing detection rules and assisting in vulnerability scanning, patch management, and policy enforcement.

Integrated security tools including Zircolite, KAPE, and Sigma to establish a robust log-based detection pipeline for compromise validation.

Contributed significantly to critical incident response workflows, collaborating directly with the CISO and Head of Cyber Defense.

Developed Python automation scripts and integrations, streamlining detection workflows and improving SIEM alert triage efficiency.

CyberXtron

|

Cyber Threat Research Intern

Chennai, Tamil Nadu, India

→

Summary

At CyberXtron, Aarav streamlined vulnerability advisories, analyzed 20+ GB of Dark Web breach data, and contributed to real-time threat detection engine updates.

Highlights

Streamlined vulnerability advisories, transforming them into actionable remediation guidance to bolster compliance and GRC functions.

Processed and analyzed over 20 GB of Dark Web breach data, generating critical vulnerability intelligence and comprehensive exposure mapping.

Delivered structured intelligence for operational use, directly contributing to real-time updates and enhancements in the core threat detection engine.

Developed Python and SQL scripts for efficient data parsing and visualization of complex breach datasets.

Education

SRMIST

Chennai, Tamil Nadu, India
→
Bachelor of Technology Computer Science Grade: SGPA of 9.6 and CGPA of 9.07

Courses

Red Hat Enterprise Linux (RHEL)

Short Range Wireless Communication Devices

Computer Networks

GPU Programming (CUDA, OpenACC, OpenCL)

Drone Analytics and Network Security

Publications

Federated Learning Aggregation for Chest X-rays: Comparing Robust Methods and Analyzing the FedHeurAgg Adaptive Heuristic

NeurIPS (submitted to)

Benchmarked six FL aggregation strategies for pneumonia detection; introduced and evaluated FedHeurAgg, a novel heuristic. Developed a novel adaptive aggregation method, improving low-heterogeneity accuracy but revealing performance limitations under high skew. Analyzed computational cost, revealing a 3-4x higher overhead in FedHeurAgg compared to baselines, identifying future optimization paths. Investigated data heterogeneity impacts, demonstrating FedTrimmedMean's superior robustness to high non-IID factors in medical datasets.

Problem-based UAV Design for Port Surveillance and Monitoring and Machine Learning for Detection and Identification of Vessels

IEEE ICECA

Co-authored an IEEE-published paper on UAV-based port surveillance and vessel detection. Proposed UAV sizing/design methodology and simulated mission architecture for real-world maritime security use cases. Developed and hyper-tuned a YOLO-v5 detection pipeline, achieving 97.8% mAP across custom datasets. Demonstrated integration with UAV simulations for autonomous encroachment detection and surveillance.

Certificates

CEHv13

EC-Council

Cyber Crime Intervention Officer – VLK5LB39

ISAC

Cybersec Foundations

Google

Deep Learning

NVIDIA

Certified Associate Cybersecurity

Fortinet

Skills

Programming and Scripting

C, C++, Python, BASH, SQL, JavaScript (basic).

Cloud & Systems

Azure, GCP, Linux, Docker, RHEL.

Cybersecurity

SIEM, Threat Hunting, Vulnerability Management, Incident Response.

AI/ML

GANS (SAR colorization), Federated Learning (medical FL), CNNs (pneumonia detection).

Projects

Deepfake Detection | HackStack National Hackathon

→

Summary

Developed a real-time deepfake detection model using transfer learning on ResNet, optimized for lightweight inference, achieving high detection accuracy.

NASA CanSat

→

Summary

Led the software team for an aerospace payload project, developing avionics systems with real-time telemetry and active fin control.

Sentrix FOSS Enterprise Security Platform

→

Summary

Architected and deployed SIEM, endpoint protection, and DLP solutions across Azure & GCP for an enterprise security platform.