Hi, I'm Ayush Tibrewal

Undergraduate at DTU, passionate about building intelligent systems with a strong interest in full-stack development, AI/ML, deep learning, and computer vision.

Explore My Work

Real-world Results

Featured Projects

QuickPick

  • AI-powered product recommender for smart shopping
  • Lets users compare, search & filter items instantly
  • Built with React, Firebase, Tailwind CSS & OpenAI API
QuickPick

TravelAI

  • Personalized travel plans based on budget, destination & trip duration
  • Saved itineraries for seamless trip management
  • Built with React, Firebase, Tailwind CSS & OAuth 2.0
TravelAI

SnapNotes

  • Auto-generates notes from lengthy PDFs
  • AI-powered Q&A for fast, context-aware information retrieval
  • Highlight, summarize & interact with documents
SnapNotes

Identity Card Checker

  • Detected Aadhar, PAN, Voter & DTU IDs with 95% accuracy
  • OCR-powered text extraction with 90%+ accuracy (Tesseract)
  • JSON-structured output for easy data handling
Identity Card Checker
More Projects

My Professional Journey

Dec 2024 - Feb 2025

Software Developer Intern at Tech Mahindra - Built backend for multi-agent network systems to automate issue detection & resolution.

Computer vision projectML model architecture
- Designed 2 multi-agent systems for automated ticketing & user communication
- Developed backend to detect network issues and verify reports
- Tech stack: PostgreSQL, Prisma, Next.js, JWT, LangGraph, GroqAPI

Oct-2023 - Mar 2024

Machine Learning Intern at Samsung Innovation Lab - Developed DL architectures for EEG-based stress classification achieving 98.73% accuracy.

Web application dashboardData visualization charts
- Engineered 3 deep learning models (ResNet18, DenseNet, VGG16) on EEG data
- Conducted SAM vs MAT dataset analysis for cognitive stress classification
- Achieved 98.73% accuracy using EfficientNetB0 on MAT dataset

Dec 2023 - Mar 2024

Research and Development Intern at Indian Institute of Technology - Delhi - Applied ML/DL on EEG data to classify cognitive load based on entropy and complexity metrics.

NLP model architectureSentiment analysis results
- Preprocessed EEG signals and conducted a literature review on entropy, fractal & complexity features
- Extracted features to identify optimal EEG channels for detecting irregularities in brain activity
- Applied ML and DL algorithms to classify cognitive load from EEG data

Aug 2023 - Sept 2023

Research Intern at IGDTUW - Built deep learning models to classify strokes from CT scans.

Microservices architectureCloud deployment
- Utilized CNNs to classify stroke vs non-stroke from CT brain images.
- Cropped 2000+ CT scans using OpenCV to isolate regions of interest.
- Achieved 98.7% accuracy using EfficientNetB0 model.

Turning ideas into impact with
Every
Click

Research Publications

Asia Pacific Conference on Innovation in Technology (APCIT) • 2024

The Effectiveness of Advance Deep Learning Architectures for Classification of Stress using Raw EEG Data

Understanding and detecting stress in today’s fast-paced world remains challenging. Previous research has explored various physiological signals, including EEG (Electroencephalography), to quantify stress. However, accurately assessing stress poses a significant challenge, particularly in preprocessing and feature extraction of EEG signals. Hence, this paper employs advanced deep learning algorithms, including ResNet18-1D, DenseNet-1D, and VGG16-1D, on raw EEG signals for automated stress classification..

Ayush Tibrewal, Shikha, Divyashikha Sethia

Image ProcessingRaw EEGDeep LearningCNN
DOI: 10.1109/APCIT62007.2024.10673521
8th International Conference on Parallel, Distributed and Grid Computing (PDGC) • 2024

Enhancing Detection Accuracy of Brain Stroke Through Cropped CT Scans With CNN Architectures

A stroke is a life-threatening condition caused by a sudden disruption in the brain's blood supply, potentially resulting in severe neurological damage or even death. Early and accurate identification of strokes is vital for prompt treatment and better health outcomes. This study utilizes advanced deep learning architectures to predict strokes and non-strokes from CT (Computed Tomography) images. Cropping techniques are employed to reduce noise and enhance image quality. Various deep learning algorithms, including EfficientNetB0, VGG16, MobileNet, and DenseNet, are explored, Notably,

Ayush Tibrewal, Priya Pahwa, Surbhi Bharti, Ashwni Kumar

Deep LearningStroke ClassificationMedical ImagingCT ScansImage Cropping
DOI: NA

Contact Me

Email Me

ayushtibrewal2004@gmail.com
Send EmailWhatsApp

Find Me Online

LinkedInGithubGScholarLeetCode