Upayan Chakraborty

Who I Am

About

uc · 2026

uc · 2025

15+

Datasets Analyzed

39+

Plant Diseases Detected

85%+

Model Accuracy

400K+

Dataset Samples

I'm a final-year Computer Science student at MCKV Institute of Engineering, passionate about turning raw data into actionable intelligence. My work spans machine learning, data pipelines, and AI-driven applications.

From building CNN-powered agriculture tools to crafting curriculum for data science workshops, I thrive at the intersection of technical precision and clear communication.

MCKV Institute of Engineering

B.Tech — Computer Science · Howrah, WB

7.34 GPA

Vivekananda Academy (CBSE)

Class 12 Science (PCM) · Shrirampur, WB

71.33%

Techno India Group Public School

Class 10 · Chinsurah, WB

83.83%

Work History

Experience

Mar 2026 – Present

AI Data Annotation Intern Remote

Utah Tech Labs · Remote

Annotating and labelling diverse data types — text, image, audio, and video — in strict accordance with project-specific guidelines to build high-quality training datasets.
Ensuring accuracy and quality of annotated datasets through systematic review and cross-validation of annotations.
Reviewing and validating annotated data to maintain consistency, resolve ambiguities, and uphold data integrity standards.
Maintaining strict confidentiality of data and project information across all assigned tasks and deliverables.
Consistently meeting productivity and quality benchmarks set by the team to support timely AI model development.

Data Annotation Quality Assurance AI Training Data

Jan 2026 – Mar 2026

Spring Intern — Advanced Data Science ISI Kolkata

IDEAS-TIH @ Indian Statistical Institute · Kolkata, West Bengal

Trained at the Institute of Data Engineering, Analytics and Science Foundation — Technology Innovation Hub at Indian Statistical Institute, Kolkata, gaining hands-on exposure to cutting-edge AI topics.
Large Language Models: Prompt engineering fundamentals, and real-world applications of LLMs across diverse domains.
Deep Learning: Neural network fundamentals, backpropagation and activation functions, and Convolutional Neural Networks (CNN).
Computer Vision: Image processing basics, image classification, and object detection overview.
Agentic AI: Autonomous AI agents, workflow automation, and practical agent design patterns.
Developed the OCEAN Personality Classification Model ↗ as the capstone project for the internship.

LLMsDeep LearningCNNs Computer VisionAgentic AI

Aug 2025 – Jan 2026

Data Analyst & ML Content Writer Remote

MacroEdtech™ · Pune, Maharashtra

Executed end-to-end data pipelines for 15+ real-world datasets, performing rigorous data preprocessing, EDA, and feature engineering to ensure model readiness.
Developed and evaluated ML models using both supervised and unsupervised learning techniques, optimizing performance metrics across diverse analytical problems.
Authored technical documentation and structured curriculum translating complex data science and ML concepts into accessible educational content for broad audiences.

Jan 2025 – Mar 2025

AI & Data Analytics Intern Top 15

Edunet Foundation × Microsoft · Bengaluru, Karnataka

Experimented with Power BI features including DAX functions, Power Query, and report automation — optimizing data transformation speed by 45%. Also worked on Microsoft Azure.
Enhanced data storytelling skills via visualization best practices, increasing user engagement with reports by 30%.
Ranked among the top 15 students of the batch across the program cohort.

Work Samples

Projects

01 — Featured

AgroCare: Plant Disease Detection & Crop Recommendation

An AI-driven agriculture solution integrating plant disease detection and smart crop recommendation. Deployed a Flask web app using CNNs (PyTorch) to identify 39+ plant diseases with treatment suggestions. Integrated IoT-based soil and environmental sensors with ML models for real-time crop recommendations. The system bridges AI with agriculture to help farmers make data-informed decisions.

PyTorch CNN Flask IoT ML Agriculture AI

OCEAN Personality Classification Model

A multi-output supervised learning framework for Big Five (OCEAN) personality classification using Random Forest ensembles (n_estimators = 200) on the IPIP-NEO 120-item dataset with 400K+ samples. Achieved ~85–90% classification accuracy across all five personality traits.

Random Forest Python scikit-learn Machine Learning

Highlights

Key Technical Achievements

Hyperparameter optimization via RandomizedSearchCV (n_iter=10, cv=3) · Comprehensive feature engineering on ~160 variables · Median imputation + StandardScaler normalization · Stratified 80:20 train-test split · Feature importance analysis for interpretability.

GridSearchCV Feature Engineering Model Eval

About

Experience

Projects

Skills

Activities

Connect With Me