Ayush Tibrewal

Aspiring Software Engineer | Machine Learning & Full-stack Developer
Delhi, IN.

About

Highly analytical and results-driven Engineering Physics student with a strong foundation in Deep Learning, Full-stack Development, and competitive programming. Proven ability to innovate and deliver high-impact technical solutions, as demonstrated by achieving 98%+ accuracy in EEG signal processing and developing AI-powered systems for network anomaly detection. Seeking to leverage expertise in machine learning, data structures, and system design to drive technological advancements in a challenging software engineering or research role.

Work

Tech Mahindra
|

Software Developer Intern

Pune, India, India

Summary

Developed an AI-powered agent and backend system for network anomaly detection, enhancing operational efficiency and data integrity.

Highlights

Engineered a LangGraph-based AI agent to autonomously detect over 100 weekly network anomalies, significantly reducing manual intervention and enabling real-time ticket generation.

Developed a robust backend system to identify network issues and verify existing reports, streamlining resolution workflows and enhancing operational efficiency.

Integrated MongoDB for high-volume network log storage and PostgreSQL for structured ticket management, eliminating duplicate entries and improving data integrity.

Samsung Innovation Lab
|

Research and Development Intern

Delhi, India, India

Summary

Designed and implemented novel deep learning architectures for raw EEG signal processing, achieving high accuracy on public datasets.

Highlights

Designed and implemented novel deep learning architectures to process raw EEG signals, advancing research in brain-computer interfaces.

Trained advanced deep learning models on public SAM40 and MAT EEG datasets, demonstrating proficiency in model optimization and data handling.

Implemented batch normalization and dropout techniques to significantly prevent overfitting, enhancing model robustness and generalization across diverse EEG data.

Achieved state-of-the-art accuracies of 98.73% (SAM40) and 97.24% (MAT) using Conv1D-based end-to-end learning for EEG signal classification.

Volunteer

Semicon DTU
|

Joint Secretary

Delhi, India, India

Summary

Organized 5+ tech events to foster student engagement in hardware, software, and semiconductor fields.

Highlights

Successfully organized over 5 technology events, engaging students in hardware, software, and semiconductor fields, fostering community and skill development.

AIMS DTU
|

Co-Head, AI-ML

Delhi, India, India

Summary

Led AI-ML initiatives for 3 major projects, including ML-integrated drones and unmanned vehicles.

Highlights

Directed AI-ML development for 3 significant projects, including the integration of machine learning into drones and unmanned vehicles, guiding project teams to successful outcomes.

Education

Delhi Technological University
Delhi, India, India

Bachelor of Technology

Engineering Physics

Grade: CGPA: 8.52

Awards

Departmental Rank

Awarded By

Delhi Technological University

Achieved 2nd rank in the department with a 9.26 SGPA across Semesters 5 and 6, consistently maintaining a top-10 standing.

OxML 2025 Summer School Selection

Awarded By

University of Oxford

Selected for the prestigious OxML 2025 Summer School by Oxford on Machine Learning, recognizing high potential in the field.

Research Internship Selection

Awarded By

Divine Lab, IIT Delhi

Selected for a competitive research internship at Divine Lab, IIT Delhi, from a pool of over 500 applicants.

HackCOG Esya'23 - 8th Rank

Awarded By

IIIT Delhi

Secured 8th rank in HackCOG, a competitive hackathon hosted by IIIT Delhi, demonstrating strong problem-solving and technical skills.

Publications

Brain Stroke Classification Using Deep Learning

Published by

8th International Conference on Parallel, Distributed and Grid Computing (PDGC 2024)

Summary

Paper accepted at the 8th International Conference on Parallel, Distributed and Grid Computing (PDGC 2024), focusing on advanced deep learning techniques (CNNs, OpenCV) for brain stroke classification.

EEG based Cognitive Load Detection

Published by

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

Summary

Paper accepted at the Asia Pacific Conference on Innovation in Technology (APCIT-2024), detailing research on EEG-based cognitive load detection using PCA and Hypothesis Testing.

Skills

Databases

MongoDB, PostgreSQL.

Cloud & Tools

Git, GitHub, Puppeteer, OAuth 2.0, Google Maps API, Location API, Image APIs, Tailwind CSS.

Core Competencies

Full-stack Development, Machine Learning, Deep Learning, Computer Vision, Data Structures & Algorithms, Object-Oriented Programming (OOPs), System Design, Competitive Programming, Backend Development, Frontend Development, AI Agent Development, Network Anomaly Detection, EEG Signal Processing, Object Detection, Itinerary Planning, Web Scraping, Data Analysis, Model Robustness, Problem Solving, Technical Leadership, Project Management.

Programming Languages

Java, Python, TypeScript, JavaScript.

Frameworks & Libraries

React, Node.js, Next.js, Express, Prisma, TensorFlow, Pandas, NumPy, JWT, OpenCV, LangGraph, Firebase.

Projects

QuickPick

Summary

A unified grocery comparison tool designed to aggregate real-time pricing and availability from major e-commerce platforms like Blinkit, Zepto, and Swiggy Instamart.

TravelAI

Summary

An AI-powered itinerary planner developed with React and Firebase, enabling users to create custom travel plans based on budget, destination, and trip length.

WalkMate

Summary

A real-time visual aid for the blind, built using TensorFlow and COCO models, designed to enhance environmental awareness and navigation.