Chinmay Pagey

Education

MTech Integrated Computer Science with specialization in Business Analytics

CGPA - 9.03

Chennai | October 2020 - June 2025

Data Analyst

Mumbai | Jan 2025 - Present

Co-leading the deployment of NexusAI to over 1,000 employees, including onboarding senior leadership and supporting enterprise-wide adoption.
Building and maintaining the NexusAI Knowledge Center, a centralized portal featuring tutorials, platform updates, and blog content to drive user engagement.
Designing and implementing AI tools using Retrieval-Augmented Generation (RAG) to enhance document search and retrieval.
contributing to the development of internal tools like HR Policy Assist and Document Query systems to improve self-service workflows.
Developing domain-specific AI agents for market research, tender analysis, and IPR automation by engineering tailored system prompts, feeding relevant datasets, and training and testing agents to ensure accuracy and alignment with business requirements.
Working with large datasets to extract insights that support data-driven decision-making across teams.

Data Analyst (Internship)

Bangalore | May 2023 - August 2023

Improved clients' marketing strategies, resulting in a 35% increase in audience reach through data-driven insights.
Engineered an automated data extraction process using Apollo CRM, enabling seamless delivery of 5000+ verified contact details to clients monthly, facilitating targeted outreach and boosting their lead conversion rates by 40%.

SEO team (Internship)

Mumbai | May 2022 - July 2022

Supported the SEO team by researching and recommending keywords for the newly launched website 'Good Health by Yourself', leading to a 400% increase in web traffic within a short period.
Gained hands-on experience with Semrush and Ahrefs to identify high-impact, relevant keywords across 9+ countries, enhancing the website's global search visibility and content relevance.

Deep Learning

Achieved 98.72% accuracy by developing OncoCutisAI, a hybrid diagnostic model combining 17 metadata features and 10,000+ augmented dermoscopic images using ensemble learning with Swin-Tiny, MobileViT, and XGBoost.
Preprocessed data for 1,373 patients, using MICE to impute 15% missing values and applying 5 augmentations per image to enhance deep learning model generalization.
Benchmarked multiple architectures (ResNet, Efficient Net, MobileNet) and ensemble techniques (majority voting, weighted averaging) to optimize performance and reliability.

Natural Language Processing

Developed an automated essay grading system using Bi-LSTM architecture, achieving 96% accuracy in grading cohesion, syntax, and vocabulary for essays by English learners (Grades 8-12).
Incorporated BERT for plagiarism detection by computing a similarity index, comparing essays to detect potential copying.

Computer Vision

Developed Nivaaran for SIH Hackathon, a smartphone app that integrates Instagram and Twitter APIs to collect real-time data on water concerns.
Applied MobileNet for flood image detection with an accuracy of 93%, enhancing its capabilities to effectively address water-related issues in communities.
Integrated the model into the app to detect flooded areas, classify water-related issues from social media images, and assess flood risk for timely disaster response.