Open to opportunities

Chandana Anugu

@chandanaanugu

Message

Data Scientist building end-to-end ML pipelines for healthcare AI, geospatial analytics, and real-time recommendations.

India

Message

What I'm looking for

I’m looking for a data science role where I can build and ship end-to-end ML pipelines, work with strong engineering practices, and apply machine learning to impactful real-world problems in healthcare and geospatial domains.

I’m a Data Science professional with an M.P.S. in Data Science from UMBC, specializing in machine learning, geospatial analytics, and healthcare AI. I build end-to-end ML pipelines—from data engineering and feature development to model deployment—focused on solving real-world problems with measurable impact.

In my projects, I forecast Chicago crime trends using ARIMA/SARIMA with 80%+ monthly prediction accuracy, and I run geospatial hotspot analysis with GeoPandas to improve patrol coverage by 50% while reducing predicted crime events by 25%. I also improved 30-day diabetic readmission prediction by 22% using ensemble models and SMOTE on 100,000+ EHR records, and I developed a streaming restaurant recommendation engine using Apache Spark and Apache Kafka to achieve 30% higher throughput and stronger ranking relevance in offline evaluation.

Experience

Work history, roles, and key accomplishments

Crime Prediction & Geo Analytics

UMBC

Jan 2025 - Present (1 year 6 months)

Built ARIMA and SARIMA time-series models to forecast Chicago crime trends, achieving 80%+ accuracy on monthly predictions and automating the end-to-end pipeline from ingestion to interactive visualization. Performed GeoPandas geospatial hotspot analysis to identify high-crime zones, resulting in a 50% increase in patrol coverage and a 25% reduction in predicted crime events.

Python ARIMA SARIMA Geopandas Time Series Analysis Feature Engineering Tableau Data Pipelines

Hospital Readmission Prediction

UMBC

Jan 2024 - Present (2 years 6 months)

Queried, cleaned, and transformed 100,000+ EHR records using SQL and Python, training ensemble models to improve 30-day readmission prediction by 22% over baseline on a held-out test set. Engineered 30+ clinical features and addressed class imbalance with SMOTE, evaluated using AUC-ROC and F1-score.

SQL Python XGBoost Random Forest Gradient Boosting SMOTE Feature Engineering F1 Score

Real-Time Restaurant Recommender

UMBC

Jan 2024 - Present (2 years 6 months)

Designed a streaming restaurant recommendation system using SQL, Apache Spark, and Apache Kafka to process geospatial and user-preference signals in real time, achieving a 30% throughput improvement over batch methods. Integrated collaborative filtering with location-aware ranking to increase offline relevance by 35%.

Apache Spark Apache Kafka SQL Collaborative Filtering Location Aware Ranking

Wireless Patient Monitoring System

CMR Institute of Technology

Jan 2023 - Present (3 years 6 months)

Developed Bluetooth 5.0 IoT patient monitoring software to collect vitals in real time, cutting data latency by 60% and achieving 92.5% detection accuracy. Implemented a critical-alert notification engine to trigger real-time alerts on anomalous readings, reducing alert latency by 40% in simulation.

IoT Python Firmware Development Monitoring Alerting

Education

Degrees, certifications, and relevant coursework

University of Maryland, Baltimore County (UMBC)

Master of Professional Studies (M.P.S.), Data Science

2023 - 2025

Grade: GPA: 4.0 / 4.0

Master of Professional Studies (M.P.S.) in Data Science from UMBC (Aug 2023–May 2025), focusing on machine learning, geospatial analytics, and healthcare AI. Completed with GPA 4.0/4.0.