Om J.

Key Skills

Software

No software listed

Top Subject Matter

Email spam detection using supervised NLP classification

Supervised house price prediction via machine learning pipeline building

Top Data Types

Text

Top Task Types

Classification

Freelancer Overview

Project: Spam Email Classifier (end-to-end text classification with FastAPI deployment). Brings 5+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include scikit-learn. Education includes Bachelor of Technology, Walchand College of Engineering (2028). AI-training focus includes data types such as Text, Computer Code, and Programming and labeling workflows including Classification, Computer Programming, and Coding.

Labeling Experience

Project: Bangalore House Price Prediction (end-to-end supervised regression pipeline)

Created an end-to-end regression pipeline for predicting residential property prices using a Bengaluru House dataset. The workflow involved cleaning and transforming the dataset into a model-ready labeled training format, followed by training and evaluating regression models to select the best performer. The final preprocessing and model steps were serialized for consistent inference during deployment. • Performed data cleaning: handled missing values, extracted BHK from text, converted sqft ranges to numeric averages, and removed implausible records • Engineered categorical location features with OneHotEncoder and standardized numerical columns with StandardScaler inside a single sklearn Pipeline • Evaluated Linear Regression, Lasso, and Ridge models and selected Ridge Regression based on R² performance • Serialized the trained pipeline with pickle to support deployment-ready inference

2024 - 2028

Project: Spam Email Classifier (end-to-end text classification with FastAPI deployment)

TextClassification

Built an SMS spam classification model by processing raw message text into features suitable for supervised learning. The work included preparing labeled training inputs from the UCI SMS Spam dataset and training Logistic Regression and Naive Bayes classifiers for binary outcomes. The trained model was evaluated with standard classification metrics to select the best approach for deployment. • Used TF-IDF vectorization to transform text into numerical representations • Compared precision, recall, and F1 across Logistic Regression and Naive Bayes • Implemented a deployed inference service with a FastAPI REST /predict endpoint • Structured training and serving code into modular components using OOP principles

2024 - 2028

Education

Work History