Lead Data Scientist & Deployment Lead | Oracle Next Education (ONE) Hackathon

• NLP Text Classification: Categorized 240,000+ financial transaction records into 10 classes with 99.0% test accuracy using Scikit-Learn TF-IDF vectorization.
• Low-Latency MLOps: Serialized Scikit-learn pipelines into ONNX format (skl2onnx) enabling native in-memory sub-millisecond inference (< 0.8ms, ↓ 98.2% latency reduction) inside a Java 17 / Spring Boot backend.
• Containerized Cloud Architecture: Orchestrated 4 microservices (Vue 3, Spring Boot, MySQL, Jupyter) using Docker Compose deployed to Oracle Cloud Infrastructure (OCI Compute).
• Zero Data Leakage: Enforced strict temporal train/test validation splits and systematic hyperparameter tuning via GridSearchCV (36 candidates, 3-fold CV).

Bilingual Static Web Platform & Interactive Latency Benchmarking Hub

• Global-First Architecture: Native bilingual routing (English default + Spanish) with zero client-side JavaScript overhead for static content.
• Interactive Visualizations: Embedded Chart.js benchmark component illustrating empirical inference latencies between Python REST APIs and native ONNX runtimes.
• Integrated ATS Resume: High-fidelity, single-page printable CV with instant PDF generation.

Applied Machine Learning & Statistical Validation Practice

• Automated Preprocessing: Modular pipeline featuring IQR outlier filtering, StandardScaler normalization, and One-Hot Encoding for large tabular datasets.
• Model Evaluation: Benchmarked baseline algorithms (Logistic Regression vs. Decision Trees) using ROC-AUC curves, Confusion Matrices, and weighted F1-Scores.

Data Science & Machine Learning

Cloud, DevOps & Databases

Web, Tooling & Frontend