PRASHANTH NAIR

Built OOP based Neural Net modelling tool using python. Built OOP based Neural Net text modelling tool using python. Built automated OOP based reporting tool using python. Built OOP based statistical package in python. Built machine learning NLP binary model with 99% accuracy. Built Neural Network NLP binary model with 98% accuracy. Built machine learning NLP multi class model with 97% accuracy. Built multi class NLP neural network model with 94% accuracy. Built Keyword Parser in Python, a generic NLP tool to extract domain specific words. Integrated the Keyword parser with single layer neural network to predict test cases with Python based W2V modeling. Built a python search engine prototype to find bugs associated with networking keywords. Built a Sckit Learn model prototype to predict components based on networking keywords.

Reconfigured and tested Audit Ensemble Modeling using R - glm, glmnet, gbm, svm and randomForest. Built models for predicting ATM and cash deposit machine (CAM) failures using GBM, this model can be scaled to analyze any mechanical failure. Built simulation queuing model in R for estimating queue length and wait time and utilization of servers. Integrated text mining models built to extract business intelligence from ATM unstructured files.

Project Churn Modeling for PLDT global Malaysian portfolio. Using Ishikawa Root cause analysis method to analyze churn. Tools used SAS 9.3 (proc sql and macros), Excel, Power point. Developed Demo's Using R programming module. Three different real world case studies using R. Module to be used for training and development using R. Comparative study of R versus SAS.

Have instructed Masters level and undergraduate classes in subject areas mentioned below in the last two years. Organized and conducted management and economics discussion group. Guided project work for undergraduate students.

Developed automated fraud detection systems using SAS. Implemented automated call schedule for outbound fraud call. Implemented Collections mailer list with SAS. Conducted Analysis and Reporting using SAS /Oracle/Unix. Developed and monitored Unix scripts for daily processes. Developed Web Based Reporting with PHP and Oracle 9i.

Developed understanding credit card business practices. Developed understanding of collection practices and risk mgmt. Monitored, reported and researched bankruptcies across chapters. Monitored and reported on consumer credit counseling services portfolio.

Developed feature engineering for AML model using R and Pyspark. Using Python(numpy, pandas, sckitlearn and pyspark and R to derive solutions for data science problems.

Developed understanding of risk in credit card Business. Monitored and reported MIS on weekly and monthly data. Researched and implemented procedure for database migration from DB2 to Teradata. Developed and recommended relationship between utilization, credit Limit and Risk.

Developed SAS based strategies and testing engine to mitigate fraud. Developed SAS based Exhaustive mining tool "FP-Miner". Developed "Fraud Dashboard". A SAS/Excel/Access based decision support system. Conceived method for identifying "True Name fraud". High dollar impact project.

Built automated call list for call operations using SQL programming. Built suggestive daily fraud generator using volume change algorithm. Recommended strategies using data mining to enhance sales. Developed complete daily operational effectiveness and monitoring program. Analyzed data patterns for strategic validation, strategic formulation and operational effectiveness.