About

Aniruddha Humane is an Applied Machine Learning Engineer at SmartNews with over 4 years of experience in developing large-scale data-driven systems. He is a data and analytics enthusiast with expertise in NLP, recommendations, market basket analysis, and optimization with distributed computing. Aniruddha has a proven track record of increasing advertisement views by up to 60% with significant increases in spend at OnlineSales.ai. He has also delivered impeccable transactional fraud detection solutions to customers all over the world at NICE Actimize. Aniruddha is proficient in programming languages such as Python, C/C++, and Java, and has hands-on experience in data science tools like Sci-kit learn, PyTorch, and Keras. He is well-versed in big data technologies like Hadoop, Spark, Pig, and Hive, and has worked with databases like Athena, Psql, MySQL, and Cassandra. He has also worked with caching and search tools like Redis, ElasticSearch, and Solr, and has experience working with Azure ML Studio, Notebooks, and Data Factory. Aniruddha has published a research paper with 7 citations and has expertise in various technologies like Azkaban, Cloudera, S3cmd, and Jira. He has a Bachelor's degree in Computer Engineering from Pune Institute of Computer Technology. Aniruddha has worked as a Data Scientist at OnlineSales.ai, where he implemented various approaches to improve ad campaign performance, performed market basket analysis at scale using the fpgrowth algorithm, and created a novel bid optimizer that maximized revenue. He also optimized the main SKU recommendation engine using Dimensionality Reduction techniques and caching to improve the recommendation quality and devised an Ad-Search terms Matcher using Approximate Nearest Neighbors and Map-Reduce operations on Hadoop, which improved the search ads response rate by up to 60%. At NICE Actimize, Aniruddha worked as a Software Engineer (R&D) and played a key role in the development of the feature that executes ML models deployed onto Kubernetes clusters through RestAPI for fraud detection. He also implemented a model explanation API using SHAP plots and led the feature that significantly reduced the time for maturation of aggregated data using historical transactions from months to just a few days. Aniruddha has provided various optimizations to the system using existing features to increase the scalability and throughput of the system during annual hackathons. These ideas later became eligible for patents. Aniruddha