Consultant Data Scientist
Consultant Data Scientist
- NLP: extraction/IR of metadata from semi-structured legal text sources (Python, Java, SQL, Postgres, pylons, CoreNLP, nltk, WordNet) for pre-IPO SaaS client.- Predictive analytics for customer retention, segment analysis (R, SQL, Python, pandas, S3, ElasticSearch, REST, Magento API, data ingestion from multiple APIs and DBs). Achieved 1-5% lift in retention for multiple US clients (retail/social startups). Actionable business recommendations.- Text analytics: sentiment analysis of reviews, POS tagging- Structured document analysis (legal & patent filings), IR/tagging, NER- Sentiment analysis in social-media post corpus: segmentation, punctuation disambiguation, lemmatization across multiple languages, Unicode encodings- Search relevance of e-Commerce user queries- Visualizations with ggplot, pyplot, (timeseries, heatmap, marimekko plot) etc.- Analysis, visualization, geomapping of datasets (US Census/ACS, zipcode, pollution)- Online Handwriting Recognition with HMM (MSEE thesis)- Tools/technologies: bag-of-words, tfidf, word2vec, CoreNLP, NLTK, WordNet API, R tm, jython, POS, Named Entity Extraction, Semantic Types Identification- ML models: RF, logistic regression, GBM, XGboost, NB, HMM, feature selection, clustering- Github: bug reporter for Python pandas, sklearn, R dplyr, data.table, committer to h2o- Coursera NLP and ML courses, UCSC-Ext ML course