Praveen Chandran                             

SUMMARY

Experienced user of data science techniques like machine-learning algorithms using SAS & Python that enables senior management to make data based decisions in media, retail & insurance industry and strong interest in deep learning(AI). 6 years of experience with 3 years of professional experience developing web applications and 3 years of experience building predictive models with SAS, Python using various machine learning algorithms like SVM, CNN, RNN, Random forest and performing various statistical tests and load data from HDFS, analyze data using HIVE, visualization/reporting tools. Main focus was on the data analytics field building predictive models using SAS Enterprise Miner, SAS Enterprise Guide & Spyder to predict possible outcome and translating the statistical results to a report understandable by the business people.

·         Adept at SAS and big data technologies like Hive, Pig, Hadoop cluster formation

·         Three years of experience in IT Consulting, database marketing and data mining using Cross Industry Standard Process (CRISP) for SAS Programming

·         Proficient in Statistical data analysis (ANOVA Tests, T-Tests, Correlations, Cluster analysis, One-way frequencies, Summary statistics) and good understanding of machine learning algorithms like Decision trees, Random forest, Logistic regression, Linear regression, Artificial Neural networks, Bagging, Gradient Boosting, SVM, Naïve Bayes

·         1.5 years of academic experience in data mining and analysis, database marketing, customer profiling, segmentation techniques, RFM, and predictive statistical modeling

·         Used SAS procedures like PROC FCMP, PROC SQL, MEANS, TABULATE, REPORT, REG, FREQ, CORR etc.

·         Used SciKit, NumPy, panda libraries in python for data cleaning and predictive modeling

·         3 years of IT experience in web and windows development using C# and writing complex SQL queries, triggers, stored procedures and functions

CERTIFICATIONS

·         SAS^® Certified Base Programmer for SAS 9 

·         SAS^® Certified Statistical Business Analyst using SAS 9: Regression and Modeling                                                                                                  

·         SAS^® and OSU Data Mining Certificate                              

·         Hortonworks Certified Pig & Hive Developer                                                                                                          

·         Cloudera Certified Hadoop Developer                                                                                                                      

·         Cloudera Certified Hadoop Administrator   

·         Machine Learning with Python – Udemy (ongoing)                                                                                               

EXPERIENCE

Deep Learning Projects[Spyder, Python]                                                                                                  [Dec’17 – Present]

Worked on the following deep learning projects as part of self-learning,

·         Built a CNN model on the CIFAR-10 dataset with 60% accuracy

·         Translated scripts from English to French using RNN.

Data Scientist at PCH, New York[SAS Enterprise Guide, SAS Enterprise Miner, Python, SQL, Hadoop]                [Dec’15 – Present]

Created various predictive models to help business target the right customers. The objective of one of the projects was to analyze internal data of offline merchandize customers who ordered retail items and join them with prospect data from outside vendor Merkle to predict the propensity of customer to buy a product or not and also predict the customer future value (CTM).

·         Pre-processing of data was done in SAS Enterprise Guide and predictive model using machine learning algorithms was done in SAS Enterprise Miner

·         The model has consistent lift value of  4.2 for the top decile and is in real time production right now making $5M in revenue

·         Created a predictive classification model with accuracy value of 0.80 in Python using SVM (Support Vector Machine) algorithm with Scikit – learn library

·         Created predictive models to deduct the data quality from two vendors, Merkle & Alliant

·         Worked with huge data (100M) to identify the right population to do the vendor analysis from PCH Hadoop database where online customers have registered through various sources

Created a SAS macro program to allocate the products in the specific mailing which is sent every week during the spring and fall season of the year

·         This project is to automate the work done by marketing and merchandize team to allocate products for each mailing using SAS program

·         Program took 6 hours to achieve what the team took 1 month to achieve the same thing

·         Won SWOT award from PCH

Data Scientist at DirecTv, Los Angeles [Pentaho, IBM DB2, Hadoop, Hive, R]                                             [Dec’14 – Dec’15]

The objective of the project is to analyze data from different vendors supplying total Movies & Series TV description, schedule, channel lineup etc. and suggest the best vendor supplying the detailed information and also accurate program schedules to business.

·         Responsible for designing and implementing ETL process using Pentaho to load data from CSV and XML files to DB2 database

·         Used SAS for post processing of data queried from DB2 database and analyze data by creating graphical and tabular reports

·         Used SAS to perform descriptive statistical analysis and analyze central tendency and dispersion measures

·         Used Crontab to automate the jobs written using Shell script for daily download of data from different vendor FTP’s and run the kettle jobs

·         Performed data analysis for large data set on Hadoop framework using Hive to do accurate study on all the datasets

Predictive Modeler at Insurance Company, USA [SAS Enterprise Miner 5.1, Base SAS, SAS EG 12.1]            [Jan’14 – Jul’14] 

A real time academic project where the client is based in Florida and the business solution which we suggested was taken         into consideration. Project was done strictly on Agile and CRISP-DM methodology.

·         Implemented machine learning techniques to predict the propensity of customers buying home owner insurance & to retain existing customers

·         Developed predictive models to enable the decisions to improve the top line for $33 MM insurance industry which was predicted to improve their business by 1%

·         Performed effective data analysis, data preparation and data transformation for a large data set

·         Performed customer segmentation and profiling to identify the likelihood of higher policy conversion

·         Analyzed the product and customer segments contribution values using segmentation and cluster analysis to understand the growth challenges it faced. Recommended targeted promotions and suggested new segments that would help increase its revenues

·         Developed several models neural network, decision tree, linear and logistic regression etc. using SAS Enterprise miner and zeroed in on the best model to increase the propensity of customers buying insurance

·         Used Tableau for visualizing and to generate reports

Software Engineer at Changepond Technologies Pvt. Ltd., India                                                                   [Nov’08 – Jul’12]

Cambridge University, UK [C# 3.5, SQL SERVER 2005, AJAX, JQUERY]

·        Developed web site for ESOL users to easily filter examiners and for examiners to apply for various levels in Cambridge University, UK. The project went live in 2010 and increased the revenue for my organization by a whopping 30% through various sub projects given to my organization by the Cambridge University

DRS Data Services, UK [C#, SQL SERVER 2005]

·         Developed a robot application for UK Government to help instructors grade the paper where various grading procedures are run automatically with the application and which saw 100% adoption of application among the instructors. This system resulted in 25% increase in repeat business and 60% increase in productivity

SWITZER Instruments [C# 3.5, .NET 3.5, Windows CE]

·        Developed windows and mobile application to read/ write RFID tags with additional functionalities like plot graph view, Encryption for date expiry for EPC writer CS 101 and CS 203 which was the first unique product from our organization to be sold to various other supply chain and electronics companies. This unique product also encouraged my organization to develop software products after this successful project

Medical Defense Union e-CHARM, UK [C#, SQL SERVER 2005]

·        Worked extensively on SQL Replication to move data from one server to another server using large number of stored procedure, triggers, views and also on front end by constantly following Rally software updated by client

Co-Owner/ Manager at Gate 7 Restaurant, India                                                                                              [Feb’12 – Jul’12]

·        Managed the restaurant efficiently while interviewing, hiring staff and increased the revenue with innovative strategies i.e. viral marketing and varying the food item by analyzing customer preferences

TECHNICAL SKILLS

Programming Languages:                            Hive, SAS, Python, Microsoft.Net - C#, VB.NET, ASP.Net

Statistical Software’s:                                   SAS Enterprise Guide, SAS Enterprise Miner, Spyder

Frameworks:                                                 Hadoop, .NET 4.0, MVC 1.0, Entity Framework

Tools & Web Development Software:     SAS Enterprise Guide and Miner, Excel, SAS, Tableau, intelliJ, Eclipse, Visual Studio 2010, NUnit, Rally, SVN

ETL Tool:                                                            Pentaho

  Databases:                                                     SQL Server 2008 and 2005, IBM DB2

Scripts:                                                          HTML, XML, Ajax, JQuery

EDUCATION

Master of Science in Data Science (Business Analytics & Machine Learning)

Oklahoma State University, Stillwater, Oklahoma                                                                                           

Bachelor of Technology, Mechanical Engineering                                                                                               

Anna University, India

OTHER ACTIVITIES