I design and ship machine learning systems that turn raw data into dependable, real-world intelligence. From computer vision for space applications at ISRO to assistive AI and healthcare analytics, I focus on robust modeling, clear evaluation, and pragmatic deployment.
I am an AI & Data Science student with solid foundations in machine learning, deep learning, and statistics. I enjoy understanding data deeply — how it is collected, where it can be noisy or biased, and how those details affect model behavior.
During my internship at ISRO, I worked on real-world AI systems for aerospace applications, including real-time object detection and analytical ML for structural behavior. These projects sharpened my focus on dataset quality, model robustness, and honest evaluation.
I’m particularly interested in computer vision, applied deep learning, and LLM-powered tools. My goal is to contribute to engineering teams where I can own well-scoped ML problems, iterate quickly, and ship solutions that are reliable and explainable.
Worked on real-time computer vision systems for aerospace-related image streams, focusing on detection accuracy, dataset quality, and low-latency inference under practical deployment constraints.
Applied machine learning techniques to structured aerospace and engineering datasets to analyze preload bolt joint behavior, emphasizing interpretability and data-driven engineering insights.
Worked as a full-stack intern contributing to web application development, focusing on clean UI implementation, backend integration, and API-driven workflows.
SSD MobileNet-based custom detector with dataset creation, training, and real-time inference optimization.
ML models for analyzing preload bolt joint behavior using aerospace and engineering datasets.
Assistive system combining facial and voice emotion recognition to support individuals with autism.
Upload PDFs or Word documents to extract structured content and generate concise, context-aware summaries using Large Language Models.
Semantic search engine for research papers with LLM-powered summarization and natural language queries.
CNN-based classification of brain MRI images to predict the presence of tumors.
Predictive analytics on health datasets with feature engineering and model comparison.
As part of my ISRO internship, I worked on a real-time object detection system using SSD MobileNet, tailored for aerospace-related image streams. The goal was to detect relevant objects with low latency under hardware and accuracy constraints.
This project focused on using machine learning to analyze preload bolt joint behavior, working with engineering and aerospace datasets. The aim was to understand and predict how structural parameters influence joint performance.
An assistive system designed to help support individuals with autism by recognizing facial and voice emotions. The system combines computer vision and audio analysis to infer emotional states.
This project focuses on using Large Language Models to automatically classify unstructured documents (such as PDFs and Word files) and generate concise, context-aware summaries. The system is designed to reduce manual review effort for documents like research papers, resumes, reports, and technical notes.
A semantic search engine for research papers that allows natural language queries and provides concise summaries powered by an LLM. The system focuses on relevance and readability for students and researchers.
A deep learning model to classify MRI images as tumor or non-tumor. While built as a student project and not for clinical use, the focus was on careful preprocessing and robust evaluation.
A supervised learning project that predicts the likelihood of diabetes using health datasets. Focused on clean preprocessing, feature engineering, and comparing multiple ML algorithms.
I'm actively looking for opportunities where I can contribute to real ML projects, learn from experienced engineers, and take ownership of well-defined problems. If you're hiring for AI/ML internships or entry-level roles, I'd be happy to share more details and relevant code samples.