Sagar Sanghavi's Email Addresses & Phone Numbers

📖 Summary

Senior Data Scientist @ Nielsen  Assisted Product Leadership in scoping, development and execution of new research plans and methodologies by trend analyses, data integration, automation, generalization, harmonization Developed new methodologies to measure cross-platform audience on new digital devices and platforms, leveraging data from multiple sources by applying statistical techniques Evaluated quality and accuracy of demographic data from these sources, using statistical models and existing Nielsen data assets  Owned analyses for the Methods group by doing product analyses, handling client inquires, scoping impact of changing methodologies, standards, and best practices  Collaborated with cross-functional teams to design, implement, and test new consumer and audience measurement methodologies  Developed code, automated methodologies and translated code from R in Python, Py-Spark and Spark-SQL San Francisco Bay AreaData Scientist Intern @ Progress DataRPM • Managed data aspects of a customer engagement by handling of huge client data greater than 10GB at times and reported discrepancies in data back to the client and came up with techniques to clean the data • Developed python script for cleaning and preprocessing of client data to get it in format consumed by data pipeline and automated it leading to decrease in effort of engineering team from 4-6 days to less 3-5 hours • Developed and deployed production level code to implement scalable supervised and unsupervised machine learning techniques in parallel processing frameworks like Spark ML or Dask infrastructure in python • Collaborated with different clients to solve their use case and deliver results for different phases within timeline• Identified anomalies on time series sensor data from different equipment data to define machine state and developed algorithm to calculate risk score and remaining usable life of the equipment to prevent failure• Developed binary search to tune the parameters of an estimator and increased the model prediction by 1.7%• Helped the company migrate to deep learning by implementing Tensor Flow and used LSTM for model training which led to an increase of 3% in the model accuracy on different client datasets which are over 1TB in size• Developed code for ensemble of scalable machine learning techniques(random forest, gradient boosted trees and so on) in Spark ML (pyspark and scala-spark) From June 2017 to December 2017 (7 months) San Francisco Bay AreaData Scientist @ Karmic Stock Broking Pvt. Ltd. • Researched stock market derivatives, hedging principles, algorithmic trading and arbitrage concepts to create a Project Requirement Document to develop hedging strategies in the option segment• Collected raw data of Nifty Futures & Options containing over 150,000 records automatically using Python script by accessing online API’s and performed data cleaning, data preprocessing, data validation and data integration • Performed data analysis and did operations using python (numpy and pandas) to understand of data• Tested different hypotheses using concepts of statistics and probability to modify initial plan iteratively• Used unsupervised (K-means, DBSCAN) and supervised learning techniques (Regression, Classification) for feature engineering and did Principal Component Analysis for dimensionality reduction of features• Deployed a Random Forest regressor algorithm and developed Binary search for hyper parameter tuning• Simulated algorithm outcomes by developing financial models like Monte Carlo for uniform probability distribution and Latin Hypercube method for joint probability distribution• Achieved an average 15% ROI/month by deploying the strategies given by algorithm on real stock market data over 3 months• Co-authored an award-winning paper “Developing Hedging Strategies in Option Segment” recognized at 2nd International Conference on Computing, Communication, Control and Automation, 2016 of IEEE) From June 2015 to May 2016 (1 year) Mumbai Area, IndiaData Analyst intern @ Sanghvi Enterprises • Designed ER model based database to store data using Microsoft SQL Server and performed descriptive, predictive and prescriptive analytics to do business planning and form business strategies leading to improvement of business process• Used Tableau for data visualization to create reports and dashboards for business insights and business development• Improved average delivery time by 4 days, increased order completion rate by 9% and revenue increase of 3%• Increased factory output by 5% by developing a system to monitor raw materials used in the factory using Advanced Excel From June 2014 to January 2015 (8 months) Mumbai Area, IndiaData Scientist Engineer @ Alpha Ori Technology • Worked with project managers of different business divisions to generate ideas on how to use data to make processes more efficient by deriving intelligence from the data• Assisted in scoping and designing analytic projects and solutions, bringing in and testing innovative ideas for out-of-box solutions and created multiple Proof of Concepts• Trained and tested the various machine learning techniques on data to investigate credence to hypotheses on different business challenges and designed algorithms do solve different business use cases• Worked with technology team to support machine learning algorithms in big data platform to solve business problems by deploying production level code in parallel processing environment• Executed analytic projects ranging from small to large both individually or as part of a project team• Developed Predictive Maintenance algorithm to decrease downtime of different industrial machines From January 2018 to October 2018 (10 months) San Francisco Bay Area