cropped-logo-1.png
Aayush Rai
Experience
Projects
Contact

Aayush Rai

Security Analyst & AI Researcher | Kansas State University

Senior at Kansas State University, double majoring in Computer Science and Mathematics (Graduating May 2026).
Specialized in AI-driven cybersecurity, threat detection, and computer vision research — with experience building enterprise security dashboards, automating intrusion detection, and developing machine learning models for real-world impact.

Resume
Contact Me

About

I’m passionate about solving complex security challenges with AI and data-driven strategies.
Currently working at Kansas State University’s Security Intelligence and Operations Center (SIOC) on projects integrating Microsoft Sentinel, Suricata IDS, and identity management solutions.

I’ve led award-winning Hackathon teams, developed ML models for predictive security and agricultural optimization, and held leadership roles in multicultural student organizations.

My focus: combining advanced technical skills with a strong problem-solving mindset to protect systems, optimize performance, and deliver scalable, real-world solutions.

Experience

Security Analyst + IAM Team

Kansas State University – Security Intelligence & Operations Center (SIOC)

May 2025 – Present

I design and deploy solutions for enterprise-level security monitoring and identity management. My work focuses on intrusion detection, dashboard integration, and automated reporting pipelines.

Key Achievements:

  • Automated weekly network scans with Suricata IDS + Nmap, generating CSV reports via jq and Bash.
  • Built centralized Microsoft Sentinel dashboard integrating Intune, Entra ID, and KQL for MDM/MAM compliance tracking.
  • Developed JMeter REST API testing framework for Grouper, enabling large-scale JSON-to-CSV auditing.

 

Tech Stack

Suricata IDS, Nmap, Bash, jq, Microsoft Sentinel, Intune, Entra ID, JMeter

Research Assistant (CNAP Project)

Kansas State University – Dept. of Computer Science

Feb 2025 – May 2025

Worked on AI models for behavioral analysis in neuroscience research, improving annotation speed and model accuracy.

Key Achievements:

  • Trained YOLOv11 and DeepLabCut pose estimation models for rat behavior recognition.
  • Optimized dataset accuracy using CVAT and Anaconda.
  • Automated preprocessing with OpenCV and JSON-CSV converters, speeding annotation by 40%.
  • Used Beocat HPC for large-scale training and evaluation.

 

Tech Stack

YOLOv11, DeepLabCut, OpenCV, CVAT, Anaconda, Beocat HPC

Research Assistant (Farmslab)

Kansas State University – Dept. of Biological & Agricultural Engineering

March 2024 – January 2025

Applied AI to precision agriculture, improving pest detection and reducing environmental impact.

Key Achievements:

  • Developed AI model to detect aphids in cereal crops, improving detection while cutting pesticide use by 80%.
  • Controlled real-time agricultural rovers via ROS/micro-ROS for pest monitoring and crop health assessment.

 

Tech Stack

Python, ROS, micro-ROS, Computer Vision, Machine Learning

Projects

ThreatHunter AI

AI-Driven Intrusion Detection Platform

August 2025 – Present

Building an end-to-end intrusion detection and explanation pipeline that combines Suricata IDS with anomaly detection and LLM-powered summaries, simulating a modern SOC automation workflow.

Highlights:

  • Integrating Suricata IDS with a Python anomaly detection model aiming to improve alert classification efficiency by 70%.
  • Developing a Fast API backend + Streamlit dashboard to visualize threats, IP activity, and alert trends in real time.
  • Adding LLM-based alert summariesfor further analyst response. 
  • Pushing enriched logs into Microsoft Sentinelvia REST APIs for SIEM alignment. 

Tech Stack

Suricata IDS, Python, FastAPI, Streamlit, LLMs (OpenAI/local), REST APIs, Microsoft Sentinel

AguaCrop

1st Place Overall – Kansas Wildcat TAPS Hackathon

October 2024

Built an irrigation optimization platform that helps farmers make smarter water management decisions using historical and forecasted data.

Highlights:

  • Designed similarity models to fill missing datasets.
  • Integrated a hyper-trained RAG chatbot for interactive insights.
  • Auto-generated and emailed dynamic PDF reports for offline access.

Tech Stack

Python, Node.js, React.js, Chart.js, Axios

Machine Learning App — Drilling Failure Prediction

Finalist – ConocoPhillips Innovation Challenge

August 2024

Developed a machine learning app to predict drilling site failures in real time using REST API data.

Highlights:

  • Achieved 81% accuracy in predicting site failures.
  • Built an ML pipeline for preprocessing, training, and evaluation.
  • Designed real-time analytics dashboard with alerting features.

Tech Stack

Python, scikit-learn, REST APIs, Data Visualization

Build Your Bowl

Point of Sale Full Stack Application

January 2024 – May 2024

Created a customizable POS system for a Chipotle-style restaurant, supporting order tracking, menu customization, and customer receipt generation.

Highlights:

  • Built full-stack system with C# backend, WPR desktop UI, and Razor Pages web app.
  • Designed fully modular UI for easy menu updates and scaling.
  • Implemented order tracking and receipt generation for smooth workflows.

 

Tech Stack

C#, WPF, Razor Pages, OOP Design, Git