Download PDF

Summary

Data Scientist and building conversational agent using Deep learning .Graduate Student in Data Science from University Of New Haven/GalvanizeU . Experience and interest in building Machine Learning, Deep Learning models. Prior experience of 10 years in Application Development and Integration of Distributed systems for large scale Investment Bank and E- commerce applications.

Education

Aug 2015Aug 2016

Masters in Data Science

University Of New Haven/GalvanizeU
Oct 2001Jun 2005

B.Tech Civil Engineering

National Institute Of Technology Jaipur

Major in Civil Engineering

Data Science Project

  • Building Dialog System and conversational platform using Deep Learning and Artificial Intelligence.
  • Building a deep learning system for chatbots which can remember long term conversation and context using RNN,LSTM , Memory networks etc
  • Build a question answer system using the facebook research dataset using memory networks.
  • Image Classification , Object Detection and Captioning system : Working on Capstone Project
  • Sentiment Analysis of Hotels using the review from Trip Advisor : Build a sentiment analysis engine using the 300k reviews from Trip Advisor.  This model was build using the NLP techniques like ngram , Word2Vec then different classification algorithm were tested like Logistic , Naive Bayes Classifier etc.
  • Real Time Mapping of geolocations in San Francisco for Uber Price Surge: This is a real time application which is built on Kafka , Spark Streaming , Hbase etc. It maps the locations in real time where price surge is happening in San Francisco it also analyses the historical pattern in price surge behavior .
  • Ranked 8th in Predicting Bidding Price Of Generic Drugs on Crowd Analytix Platform.

Experience

May 2016Present

Deep Learning Research

Msg.AI - Artificial Intelligence for Conversational Commerce
  • Building a deep learning system with memory using neural network to solve the shortcoming of understanding long term conversation and context for conversational agent.
  • Text classification system using RNN-LSTM network.
  • Trained domain specific corpus using word2vec and skipgram model.

Mar 2016August 2016

Graduate Student Researcher

Fox Type - Artificial Intelligence for writing smarter emails.
  • Image Classification , Object Detection and Captioning system : Understanding different architectures like VGG- Net , ResNet and then building a system which can build a natural description of images. This system will be built using Deep Learning models i.e CNN(Convolutional Neural Netwrok) and RNN(Recurrent Neural Network)
May 2013Jul 2015

Analyst

Barclays Investment Bank
  • Design and Development of Regulatory Reporting Framework:
  • Automation of Build and Deployment process of Tibco BW and Business Event application reducing the number of resources and effort significantly. Scaled across multiple applications.
  • Design and development of Rule Engine using Jboss Drools.
  • Mentoring team mates , co-ordination with different teams based in New York , London etc for requirement gathering , architecture design etc.

Jun 2011May 2013

Senior Associate

Sapient

JC Penney Feeds and Events:

  • Integration of multiple system into e-commerce web channel i.e.Pricing, Catalogue, Inventory, Account Profile etc. Catalog Integration into JC Penney ecommerce channel. 
  • Implementation of Catalog Integration into JC Penney ecommerce channel. With Catalog Integration one can view the PP, LOT,and Swatch level details at Browse site.
  • Migration of entire data from source JCP1 to the new ATG BCC repository.
  • Design and Implementation of Scene 7 Integration. This helps in viewing the dynamic images at Browse site.
  • Creating Load Balanced and Fault Tolerant EMS servers for high throughput and low latency.

Target  :

  • Setting up Middleware Infrastructure for Target using IBM MQ and MB product suit.

EISS(Enterprise Integration Shared Services): 

  • Building of Web Services and SOA platform for EISS team.
Jul 2007Apr 2011

Senior Software Engineer-Tech Lead 

Citigroup
  • Senior Developer with responsibilities like requirement gathering, development, end to end testing and Support to the client across the globe.
  • Design and Development of Global Confirms System : Middleware application which sends out confirms to different institutional clients through different mediums like Swift,OG,Email,Fax-Telex.
  • Development of different workflows using Tibco BW and Fuego BPM.

 

Sep 2005Jun 2007

Programmer Analyst

Cognizant
  • JP Morgan Chase - Global Commodities and Currency Group: Development of Risk and Reporting system.
  • Development of Enrichment Functions,Polling and reporting framework for different Front Offices.

Skills

Technical Skills:

Building Supervised and Unsupervised Machine Learning Models, Predictive Analytics, Deep Learning, Service Oriented Architecture , Webservice etc

Machine /Deep Learning

Building deep learning and machine learning models eg: Linear Regression, Logistic Classifier , SVM , Decision Trees , Random Forest , Ensemble Modelling, Convolutional Neural Network, Recurrent Neural Network etc

Programming Languages and Libraries

Python, Java, Tensorflow,Ant ,Pandas,Numpy,Scikitlearn,Graphlab,Keras

Tools

Spark , Kafka, Spark Streaming , Hbase ,Hadoop, Eclipse, TIbco BW , Spring Btach, Spring Integration,Drools, Fuego BPM tool

Awards

  • Best Performer of the Quarter at Citigroup
  • Client Focused Delivery Award at Sapient
  • Cognizant Certified Professional in Core Java
  • Amongst the top .1% students in India in Mathematics in class X.