Open to opportunities

Rahul Bolloju

@rahulbolloju

Message

I build reliable machine learning and experimentation systems that improve operational decisions.

United States

Message

What I'm looking for

I'm looking to build measurable data science and machine learning solutions where rigorous experimentation, reproducible evaluation, and clear communication improve real-world outcomes.

At Pitney Bowes, I build Python and SQL data pipelines across 1.2 million shipment records and develop XGBoost cost-prediction models with stable 8.2% MAPE. I also designed A/B testing frameworks that validated an 11.5% shipping-cost reduction in randomized trials.

I monitor production model performance and AI inference latency for real-time shipment scoring, maintaining 99.5% availability and 420ms 99th-percentile latency. My work emphasizes reproducible evaluation, evidence-based grading criteria, and clear reporting on KPI changes and treatment effects.

Previously at Digital Nirvana, I improved signal-analysis accuracy by 20%, reduced false positives by 12%, and automated nightly data-quality tracking pipelines. I bring rigorous statistical modeling, independent execution, and iterative feedback into machine learning evaluation and deployment.

Experience

Work history, roles, and key accomplishments

Current

Data Scientist

Current

Pitney Bowes

Jul 2023 - Present (3 years 1 month)

Translated business requirements into scalable Python and SQL pipelines, ensuring data accuracy across 1.2 million shipment records. Developed machine learning models with XGBoost achieving 8.2% MAPE and designed A/B testing frameworks that validated an 11.5% shipping cost reduction.