Riya D.

Key Skills

Software

Other

Top Subject Matter

Investment forecasting

Financial modeling

Real estate sector

Top Data Types

Text

Top Task Types

Data Collection

Text Generation

Freelancer Overview

I've worked with real-world Airbnb data by cleaning, organizing, and transforming listings, reviews, calendar availability, prices, ratings, and neighborhood data in MongoDB. This gave me experience with data quality, inconsistent formats, missing values, schema design, and preparing structured data for analysis. My background in computer science helps me work carefully with complex information and follow detailed guidelines. I am comfortable reviewing data for accuracy, identifying inconsistencies, and explaining patterns, which are key skills for data labeling and AI training data work.

Labeling Experience

Airbnb for MongoDB Prototype

TextClassification

I worked with a real-world Airbnb dataset containing listings, reviews, calendar availability, prices, ratings, room types, and neighborhood information. I helped clean, structure, and organize the data for a MongoDB-based prototype, including converting inconsistent field formats, handling missing values, standardizing city and neighborhood labels, and preparing the data for query-based analysis. The project involved building a data pipeline and a frontend prototype that allowed users to explore Airbnb records and run MongoDB aggregation queries. My work required attention to data quality, schema design, consistency across related records, and accurate interpretation of query results. This experience strengthened skills relevant to AI training data work, including data validation, categorization, annotation-style organization, error checking, and the preparation of structured datasets for downstream analysis.

2026 - 2026

SMS Spam Classification/AI Training Data Evaluation

TextClassification

Developed and evaluated an AI text-classification pipeline for SMS spam detection using labeled training data. Worked with a dataset of SMS messages annotated as either "spam" or "ham," performing data preprocessing, label validation, feature engineering, and model evaluation. Applied TF-IDF vectorization to convert text into machine-learning features and trained multiple classification models, including Logistic Regression and Multinomial Naive Bayes. Evaluated model performance using accuracy, precision, recall, F1-score, ROC-AUC, confusion matrices, and cross-validation. This project strengthened skills in data annotation quality assessment, text categorization, NLP preprocessing, model benchmarking, and AI evaluation workflows. Experience included identifying labeling patterns, analyzing classification errors, validating training data consistency, and comparing baseline rule-based approaches against machine-learning models. The project demonstrated the ability to work with labeled datasets, generate reliable evaluation metrics, and contribute to AI training data pipelines for supervised learning systems.

2025 - 2025

Education

Work History