SUNDARAKRISHNAN N

AI/ML Engineer | Software Developer

Bengaluru, IN.

About

Highly motivated Computer Science student with hands-on experience in AI/ML model development, real-time systems, and scalable software solutions. Proven ability to drive significant improvements in detection accuracy and update frequency, applying expertise in deep learning frameworks and large-scale data processing. Eager to leverage strong technical skills and leadership experience to contribute to innovative software and AI-driven projects.

Work

Zscaler

|

Software Development Intern

Bengaluru, Karnataka, India

→

Summary

Developed Agentic AI models and scalable ML pipelines for malware threat detection, significantly enhancing system efficiency and accuracy.

Highlights

Engineered Agentic AI models with Gemini 2.5 Pro, automating malware-specific threat detection and enabling monthly updates via scalable ML pipelines leveraging Kubeflow, Docker, and BigQuery for 5+ PB security data analysis.

Improved update frequency by 6x and enhanced detection accuracy, significantly accelerating response to evolving cyber threats.

Samsung Research Institute

|

Project Intern

Bengaluru, Karnataka, India

→

Summary

Developed a hair strand segmentation pipeline and a real-time hair animation system, leveraging advanced image processing and dynamics.

Highlights

Designed and implemented a hair strand segmentation pipeline utilizing multi-scale Gaussian filtering and Sobel edge detection to accurately isolate individual strands under different lighting and hairstyle conditions.

Constructed a real-time hair animation system using position-based dynamics and facial landmark detection, achieving a minimal short-term warping error of 0.083 for wind-responsive cinemagraphs.

Volunteer

Machine Learning Lead

|

Astra Robotics, RVCE

→

Led machine learning initiatives and robotic applications, including conducting workshops for students.

Highlights

Led Machine Learning and OpenCV initiatives for various robotic applications at Astra Robotics, RVCE.

Conducted an Embedded C workshop for over 250+ school students across 5 schools, fostering technical education in embedded systems.

Web Team Lead

|

IEEE-RVCE

→

Coordinated and led web projects for IEEE-RVCE, ensuring successful delivery of web-based initiatives.

Highlights

Coordinated and led the entire Web Team at IEEE-RVCE, overseeing web projects built with SpringBoot, Node.js, and React.

Managed project timelines and resource allocation to ensure timely and successful completion of web-based initiatives.

Education

RV College of Engineering

Bengaluru, Karnataka, India
→
BE Computer Science Grade: CGPA: 9.11

Publications

Vehicle Turn Pattern Counting and Short Term Forecasting Using Deep Learning for Urban Traffic Management System

IEEE Access Q1 Journal

Developed a real-time solution for predicting vehicle turn patterns using a custom YOLOv8 object tracker and forecasted traffic 30 minutes ahead with Auto ARIMA on challenging Indian roads. Secured first place in Phase 1 of the Bengaluru Mobility Challenge held by IISc, competing against over 400 teams.

YOLOv8: A Novel Vehicle Mirror Detection Algorithm with Enhanced Performance using Deep Learning

ICEARS IEEE 2025 Conference

Developed a robust vehicle mirror detection model using the deep learning framework YOLOv8, achieving a precision of 0.94 and an F1-score of 0.91. This first research paper addressed a specific hazard and involved collaboration with Bengaluru Traffic Police, utilizing real-time feed from 50 AI cameras to improve accuracy.

Languages

English

Skills

Programming Languages

Python, Java, C, Node.js.

Tools and Frameworks

Java SpringBoot, Django, Tensorflow, PyTorch, Kubernetes, BigQuery, Flask, Unity, Docker, Next.Js, React, OpenCV, JPA, MySQL, Oracle Cloud.

Soft Skills

Effective Communication, Leadership, Critical Thinking, Project Management, Team Coordination, Mentorship.

Interests

Chess

Competitive Chess Player, FIDE Rating 1456.

Projects

Intelligent DNS DGA Defense System

→

Summary

Developed a high-performance local DNS resolver for DGA detection.

GateKeeper

→

Summary

Developed an approval system for IEEE-RVCE event planning and permissions.