Harsh Raj

Computer Science Engineer specializing in Machine Learning, Deep Learning, and Generative AI
Bhopal, IN.

About

Highly motivated Computer Science Engineer specializing in Machine Learning, Deep Learning, and Generative AI, with a strong academic foundation and practical experience. Proven ability to deliver impactful solutions, including a fraud detection system achieving 92% accuracy and a multimodal house price estimator with an R2 score of 0.92. Recognized in competitive hackathons and actively contributes to open-source projects, demonstrating a commitment to applying advanced AI/ML techniques to solve real-world challenges.

Work

Visyntro
|

Data Intern

Remote

Summary

Managed client lead generation and automated data workflows to enhance reporting accuracy and project tracking.

Highlights

Generated and nurtured client leads while coordinating communication between clients and internal teams to streamline operations.

Automated ETL workflows using Python and Google Sheets API, improving reporting accuracy and project tracking by 35%.

Social Summer of Code
|

Open-Source Contributor

Remote

Summary

Actively contributed to open-source projects, enhancing code quality and community engagement.

Highlights

Collaborated with 8000+ developers on open-source projects hosted on GitHub during SSOC, demonstrating strong teamwork and version control skills.

Merged 4 pull requests by refactoring code, fixing issues, and integrating new features, improving software reliability and functionality.

Revamped contributor onboarding guides, increasing repository engagement by 25% and fostering a more inclusive community.

Vroom Verse
|

UI/UX Designer

Remote

Summary

Designed and deployed responsive user interfaces, significantly enhancing user experience and engagement.

Highlights

Crafted and deployed a responsive UI using Figma and Tailwind CSS, reducing page load times by 40%.

Refined user interaction flows and improved visual hierarchy, elevating user engagement by 25%.

Volunteer

VIT Bhopal University
|

Core Member - Data Science Club & Dance Club

Bhopal, Madhya Pradesh, India

Summary

Led technical workshops and coordinated inter-college events, fostering community growth and engagement.

Highlights

Spearheaded over 5 technical workshops, growing participant turnout by 60% through engaging content and effective promotion.

Mentored juniors and coordinated inter-college technical and cultural festivals, enhancing student development and community spirit.

Education

VIT Bhopal University
Bhopal, Madhya Pradesh, India

B.Tech

Computer Science and Engineering

Grade: 8.48/10 CGPA

Awards

Semi-finalist: Karnataka State Police Datathon

Awarded By

Karnataka State Police

Achieved a semi-finalist position among the top 6 of 250+ teams by solving traffic congestion and encroachment analysis using YOLO and OpenCV, showcasing expertise in computer vision and data analysis for public safety.

Finalist: Hackstasy Hackathon

Awarded By

SRM University (Microsoft + YouTube)

Achieved a Top 10 finalist position among 500+ participants by devising a real-time traffic signal optimizer leveraging OpenCV and ML-based congestion heatmaps, demonstrating advanced problem-solving and technical application.

Certificates

AWS Academy Cloud Architect Associate Level

Issued By

AWS Academy

MERN Stack Full-Stack Development Certification

Issued By

Online Platform

Web Development Specialization

Issued By

Coursera

Skills

Programming Languages

Python, C++, SQL.

ML/DL Frameworks

TensorFlow, Scikit-learn, OpenCV, PyTorch.

Web Development

Flask, RESTAPIs, Figma, Tailwind CSS, MERN Stack.

Development Tools

Git, Docker, GitHub, Google Colab, Version Control.

Cloud & Data Platforms

AWS, ETL Pipelines, Google Sheets API, PostgreSQL, Mage AI.

AI/ML Concepts

Machine Learning, Deep Learning, Natural Language Processing (NLP), Generative AI, Anomaly Detection.

Computer Vision

OpenCV, YOLO.

Projects

Fraud Detection using GenAI

Summary

Developed an anomaly detection system to identify fraudulent and irregular expense claims, achieving 92% accuracy.

Multi-Modal House Price Estimation

Summary

Architected a multimodal regression model utilizing 1500+ property listings with both textual and image data.