Agrim Rai

Agrim Rai

Computer Science Engineering Student

Email: agrim@prodijee.in Phone: +91 9896228016 Website: rai.is-a.dev GitHub: github.com/agrim-rai LinkedIn: linkedin.com/in/agrim-rai

Summary

Computer Science Engineering student at NSUT Delhi specializing in Data Science. AI Research Intern with experience in machine learning optimization and deployment. Built platforms serving 200K+ users. JEE Main 2024: 99.015 percentile (Top 0.9% among 1.7M+ candidates).

Education

Netaji Subhas University of Technology

Delhi, India

B.Tech, Computer Science & Engineering (Specialization: Data Science) | 2024 – Present

  • Relevant Coursework: Data Structures & Algorithms, Linear Algebra, Calculus, Probability & Statistics, Discrete Mathematics, Machine & Deep Learning.
  • Key Achievement: Secured an All-India Rank of 15,722 in the JEE Main 2024 exam (99.015 Percentile), placing in the top 0.9% among over 1.7 million candidates.

St. Francis de Sales School

Delhi, India

Class X & XII (CBSE Board) | Aggregate Score: 90% | Completed: March 2023

Projects

ProdiJEE.in prodijee.in

Developed a high-traffic web platform to predict scores and percentiles for the JEE Mains examination, India's largest engineering entrance test with over 1.7 million candidates annually.

  • Attracted and served 200,000 unique users and processed 37,000 complete exam response sheets within 10 days of launch.
  • Delivered highly accurate percentile ML based predictions (±0.5 error margin) for 75% of users, releasing results a full 7 days before the official government announcement.
  • Implemented advanced machine learning algorithms to analyze exam patterns and predict percentile distributions across multiple exam shifts.
  • Built scalable backend infrastructure to handle massive concurrent user traffic during result prediction periods.

Ipu.life

Developed ipu.face, a comprehensive social and academic analytics platform for 30+ colleges under IPU University, Delhi, serving a community of 225,000 students.

  • Engineered a large-scale facial attractiveness rating system leveraging a dataset of 183,000 student images, enabling peer-based attractiveness and similarity scoring.
  • Integrated college transcript result analytics, allowing students to compare academic performance and receive personalized insights.
  • Implemented friend-based rating features, letting users rate and view attractiveness scores within their social circles.
  • Created facial embeddings and a "find similar faces" tool using deep learning, facilitating discovery of lookalikes across the university network.
  • Ensured privacy and security by anonymizing sensitive data and providing opt-in controls for all rating and analytics features.

Work Experience

CV and AI Intern - Big Vision LLC

May 2025 - August 2025

  • Enhanced the functionality of Tutorji.in, an AI doubt solving assistant serving over 20k users, by implementing new features for the web application and the Chrome extension.
  • Designed and implemented a secure monetization system, creating a paywall that introduced premium subscription tiers and controlled feature access.
  • Partnered with clients to develop and deploy bespoke AI solutions, successfully delivering projects that met specific real-world business requirements.

Skills

Languages: Python, C++, SQL, MATLAB

ML/DL Frameworks: PyTorch, TensorFlow, Keras, Scikit-learn, Hugging Face, JAX

Computer Vision: OpenCV, Pillow, Torchvision, CNNs, Vision Transformers (ViT), Object Detection (YOLO, Faster R-CNN), Semantic Segmentation, Generative Models (GANs, Diffusion)