Aniruddha Madurkar

About

Aniruddha Madurkar is a technical writer with a background in data science. He specializes in writing on data science, machine learning, and AI topics. Aniruddha is also a contributor to the data science publication, Towards Data Science. As a senior data scientist, he builds end-to-end enterprise machine learning systems for Fortune 500s and startups. At Fulcrum Analytics, he worked on evolving customer segmentation model for a healthcare and insurance provider to include entire business units, greater explainability, and faster inference.

Employment

5 roles

Dec 2021 - Present 4 years 5 months

Senior Data Scientist

Fulcrum Analytics

Worked on evolving customer segmentation model for a top 10 healthcare and insurance provider to include entire business units, greater explainability, and faster inference.
Built end-to-end machine learning systems for top companies and startups across financial services, healthcare, and retail sectors.

Jul 2021 - Present 4 years 10 months

Technical Writer

Medium

Wrote on data science, machine learning, and AI topics.
Contributed to the data science publication, Towards Data Science.
Gained over 200 followers in six months of writing. Average 10K views monthly on Medium.

May 2021 - Present 5 years

Wildlife Conservation Research Assistant

University of Michigan

Led a team of three to build web-based explainable machine learning applications that assist law enforcement officers and researchers to proactively combat Wildlife Trafficking via imports into the United States.
Worked on peer-reviewed publication.
Used Gradient Boosted Trees (Supervised Learning), we predicted action/disposition from over 1 million shipments to identify packages of concern and highlight feature importance for targeted investigations.
Used K-Prototypes (Unsupervised Learning) and word embeddings, we constructed word clouds composed of keywords to highlight shipment contents at scale.

Mar 2021 - Dec 2021 9 months

Data Scientist/Engineer

Jackson National Life

Built Long Term Forecasting time series model to guide Operations’ monthly and yearly budget decisions.
Used Natural Language Processing and Dimensionality Reduction on Advisor/Client data to improve the Sales Forecasting model in production. Through model monitoring, we found the new model far more robust to data drift.
Built production-ready automated data pipelines on Azure via Databricks that have saved the Operations Analytics team 20 hours/month.
Measured a suite of KPIs for Operations using Bayesian estimation to quantify uncertainty and guide budget decisions and resource allocation. Visualized and presented results in Tableau to executive team.
Led the enterprise-wide Data Science and Artificial Intelligence Community of Practice.

Sep 2017 - Mar 2021 3 years 6 months

IT Metrics & Forecasting Analyst II

Jackson National Life

Used Python and Tableau to explore large datasets and extract insights for IT executive leaders.
Worked on insights ranging from understanding the distributions of system outages and failures, finding outliers and antipatterns during an Agile Transformation, calculating trends and variance differentials in Budget and Actual spend, and highlighting gaps in data quality.