Desire E.

Key Skills

Software

Dataloop

Encord

Label Studio

V7 Labs

Snorkel AI

CVAT

Labelbox

SuperAnnotate

Prodigy

Other

Top Subject Matter

Fintech and Finance

Customer/household segmentation using clustering (ML)

Top Data Types

Computer Code Programming

Document

Text

Top Task Types

Computer Programming/Coding

Data Collection

Transcription

Classification

Freelancer Overview

NLP Sentiment Analysis project (sentiment classification; model training and evaluation). Core strengths include Other. Education includes Online Course, Saylor University and Online Course, WorldQuant University. AI-training focus includes data types such as Computer Code and Programming and labeling workflows including Evaluation, Rating, and Classification.

Labeling Experience

Churn Prediction Application (training ML models and generating explanations)

OtherClassification

Developed a customer churn prediction platform using supervised machine learning. Implemented XGBoost for predictive modeling and SHAP for churn-driver interpretation. Delivered outputs that support telecom providers in identifying at-risk customers and explaining model signals.•Prepared structured customer features for churn modeling.•Trained an XGBoost classifier for churn risk prediction.•Generated SHAP explanations to interpret churn drivers.•Integrated backend prediction and explanation APIs for downstream use.

2023 - 2023

NLP Sentiment Analysis project (sentiment classification; model training and evaluation)

Other

Built and evaluated NLP sentiment models that classify customer reviews as positive or negative, using TF-IDF feature engineering. Compared traditional machine learning approaches with a BERT transformer model for performance. Implemented the end-to-end workflow from text preprocessing through model inference for practical review classification.•Collected and prepared text review inputs for model training/testing.•Engineered TF-IDF features from raw text.•Trained and benchmarked multiple text classification models.•Performed model evaluation to compare BERT versus traditional approaches.

2023 - 2023

Customer Segmentation Dashboard (Dash) - K-Means + PCA segmentation

Other

Created a customer segmentation dashboard using unsupervised learning to group U.S. households. Applied K-Means clustering along with PCA for dimensionality reduction and feature visualization. Packaged the results into an interactive Dash web application to present discovered financial patterns.•Prepared and cleaned the 2019 Survey of Consumer Finances dataset for clustering.•Engineered features suitable for K-Means segmentation.•Used PCA to reduce dimensionality and support analysis.•Visualized clustering outcomes in a user-facing interactive dashboard.

2022 - 2022

Education