Skip to content
View kdeepr's full-sized avatar
0 followers · 3 following

Block or report kdeepr

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
kdeepr/README.md

Hi, I'm Kuladeep Roy 👋

Typing SVG

Portfolio LinkedIn Email

About Me

Data Science & Analytics graduate student with ~3 years of industry experience spanning IT infrastructure and manufacturing quality data. I made the leap from Mechanical Engineering → Data Science, bringing a systems-thinking mindset to ML modeling, statistical analysis, and data pipelines. Currently building an AI-powered assistant (n8n + Google Gemini) for GVSU's Language Resource Center.

  • 🎓 M.S. Data Science & Analytics — Grand Valley State University, Michigan (Expected Dec 2026) · GPA: 3.68
  • 🏭 B.Tech Mechanical Engineering — Vel Tech University, India
  • 💼 Prior experience: Infrastructure Analyst @ DXC Technology · Data & Quality Intern @ Isuzu Motors
  • 🌍 Domain experience: IT Services · Automotive Manufacturing · Higher Education
  • 🔍 Seeking: Fall 2026 Data Analytics / Data Science Internship (Grand Rapids, MI · open to remote)

Tech Stack

Languages & Core

Python R SQL

ML & Statistics

scikit-learn PyTorch LightGBM tidymodels GLM FAISS

Data & Cloud

Pandas NumPy MySQL MongoDB Databricks AWS

AI Agents & Tools

n8n Google Gemini Power BI Streamlit Docker Git Quarto

Certifications

Badge Certification Issuer Date
AWS AWS Certified Cloud Practitioner Amazon Web Services YOUR_DATE_HERE

Featured Projects

Project Description Stack
Soil-Moisture-Predictive-Modeling 93-day recursive time-series forecast using delta modeling and an inverse-RMSE weighted 5-model ensemble (Ridge, RF, Extra Trees, GB, LightGBM) with TimeSeriesSplit validation Python · scikit-learn · LightGBM
AI-Powered-Image-Retrieval Content-based image search over 2,020 images — fine-tuned ResNet18 embeddings + FAISS vector database for millisecond retrieval via a Streamlit app PyTorch · FAISS · Streamlit
Predicting-K-12-Success Logistic, LASSO, LDA, Poisson & multinomial GLMs on 15,500-student DoE survey data quantifying how school choice and family engagement predict success (ROC-AUC > 0.70) R · tidymodels · GLM
US-Gun-Violence-Analysis Statistical analysis of 2014–2023 incidents merged with Census data — permutation testing, bootstrap CIs, and interactive Leaflet mapping R · Quarto · tidyverse

GitHub Stats

GitHub Stats Top Languages

GitHub Streak

Highlights

  • 📈 Kaggle Soil Moisture Challenge — 0.00033 RMSE with a custom seasonal-offset ensemble and recursive 93-day forecasting
  • 🤖 Built and deployed a production AI chatbot for GVSU's Language Resource Center (n8n + Gemini 2.5 + Google Sheets)
  • ☁️ AWS Certified Cloud Practitioner

"In God we trust. All others must bring data." — W. Edwards Deming

Profile Views

Pinned Loading

  1. US-gun-violence-data-analysis-in-R US-gun-violence-data-analysis-in-R Public

    Statistical analysis of US gun violence incidents (2014–2023) exploring temporal trends, geographic patterns, and socioeconomic factors using R and Quarto.

  2. Predicting-k-12-Success-using-generalized-linear-models Predicting-k-12-Success-using-generalized-linear-models Public

    Built GLMs in R using U.S. Census education data to analyze socioeconomic factors affecting student performance across school districts. Applied cross-validation and AIC-based model selection, and …

  3. AI-Powered-Image-Retrieval-Vector-Database AI-Powered-Image-Retrieval-Vector-Database Public

    Content-based image retrieval system using ResNet18 embeddings and FAISS vector database with Streamlit interface.

    Python

  4. Soil-Moisture-Predictive-Modeling Soil-Moisture-Predictive-Modeling Public

    93-day soil moisture forecasting using delta modeling and inverse-RMSE weighted ensemble of Ridge, Random Forest, Extra Trees, Gradient Boosting and LightGBM with 5-fold TimeSeriesSplit validation.

    Jupyter Notebook

  5. Language-Resource-center-AI-Assistant- Language-Resource-center-AI-Assistant- Public

    AI-powered front-desk assistant for GVSU's Language Resource Center, built with n8n, Google Gemini, and Google Sheets.

  6. -Exploratory-Data-Analysis-on-Layoffs-and-Impact-on-AI -Exploratory-Data-Analysis-on-Layoffs-and-Impact-on-AI Public

    Forked from Sreyaakm/-Exploratory-Data-Analysis-on-Layoffs-and-Impact-on-AI

    Applied exploratory data analysis to examine worldwide layoffs and the effects of AI on workforce patterns.

    Jupyter Notebook