I am a full stack data scientist, harvesting data from its origin in raw form to finished end product in form of insights and strategy. I am generally sandwiched between Business Strategists and Marketeers, from problem definition to insight sharing, on one side and Programmers and Data Engineers, from prototyping to deployment, on the other side. Be it conventional statistics or modern data mining algorithms, I apply what is necessary to solve the problem.
I have worked on developing several commercially deployable machine learning models such as automated risk underwriting for an insurance company, financial product recommender system for a bank, airtime transfer propensity predictor for a telecom company etc.
I use R, Python, PostGres, HADOOP, Spark, Tableau, Qlik, D3.js etc. I work end to end from Data Engineering to Data Analysis to Data Visualization.
I teach Quantitative Research and Advanced Multivariate Methods course to PhDs and regularly conduct training workshops on Tensorflow, Keras, Theano, CNN, RNN, dplyr, rstanarm, knitR, ggplot2, Pandas, NumPy, SciPy, Bokeh, Matplotlib, Seaborn, Tableau, Qlikview, D3.js and Scikit learn regularly.