Software Developer Data

ABHISHEK BOSE

****.*@*****.***.*** • 857-***-**** • http://linkedin.com/in/abhishek-bose-cs • https://github.com/abhishek-bose-cs EDUCATION

Northeastern University, Boston, MA Expected-May 2020 Master of Science in Data Science, GPA – 3.9/4.0

Relevant Courses: Supervised Machine Learning, Unsupervised Machine Learning, Data Visualization, Business Intelligence Visvesvaraya Technological University, Bangalore, Karnataka, India Sep 2006 - June 2010 Bachelor of Engineering in Computer Science, GPA – 3.6/4.0 TECHNICAL SKILLS

Programming Languages: Python, R, PL/SQL, Java, J2EE, C, C++, Scala, MATLAB, JavaScript, XML, HTML5, CSS Libraries Tools and and IDE: Frameworks: TensorFlow, AWS, PyCharm, Theano, IntelliJ, PyTorch, Tableau, Scikit-Power Learn, BI, MS Keras, Excel, Pandas, SaaS, Sisense, NumPy, Vertica, PySpark, Cubes, Flask Qlik ML Algorithms: CNN, RNN, SVM, LSTM, Random Forest, Gradient Boosting, NLP, KNN, SGD, PCA, LDA Data Warehouse: MySQL, PostgreSQL, MongoDB, EXPERIENCE Hadoop, Kafka, HQL, Snowflake, SQL Server, Redshift Data Science Co-op at Gethr, Boston, MA Sep 2019 - Dec 2019

• Streamlined data processing pipeline using AWS – Elastic Map Reduce and Spark to capture real-time user events

• Automated subscriber activities by scheduling data processing clusters between user application and DynamoDB

• Performed data cleaning on ~1 TB dataset and formulated statistical analysis required for result reporting in Tableau

• Modeled data using AWS Sage Maker to provide sales quote on offered services based on customer demographics

• Achieved 96% model accuracy by altering the hyperparameters which was controlling the learning process Senior Software Developer at IBM, Bangalore, India Aug 2014-Aug 2018

• Aggregated variety of Excel spreadsheets for data into a Data Lake and performed churn analysis through CRM data

• Visualized significant insights, built Random Forest classification model in Scikit-learn and pushed result into web API

• Developed complex SQL Procedures for Siebel Order Management and integrated them with Batch Framework

• Performed extensive EDA on asset management using PowerBI to optimize overall operational process by 12% Software Developer at Accenture Consulting, Bangalore, India Nov 2010-Aug 2014

• Worked as SQL developer to redesign database by enhancing queries, triggers and implemented replication

• Optimized code using indexes, views and stored procedures to achieve 20% better database performance

• Improved the scalability of the application by 40% by converting SOAP API to REST API for better synchronization

• Developed custom user role-based multifunctional Credit Monitoring Dashboard on Oracle Business Intelligence (OBI)

• Designed the custom dashboard with the support of Data Warehouse Engine and Oracle Data Integrator (ODI)

• Enabled multi-entity functionality by integrating Oracle Banking Platform (OBP) and OCH using XML, JSON and Siebel ACADEMIC PROJECTS

Twitter Data Analysis Python, Spark, AWS EMR, Tableau

• Extracted and streamlined Twitter data using Apache Spark along with AWS EMR to generate RDD in Python

• Implemented text processing and sentiment analysis to create a model to identify most positive/negative trending tweets over specific period of time

Analyzing the Place of Business check-in in Facebook Python, Gradient Boosting, K-nearest neighbor

• Designed and developed Gradient Boosting and k-nearest neighbor model for predicting place of business based on Facebook check-in history and footfall for ~30 million data from Facebook API

• Improved model accuracy by validating ROC AUC and F1-score to achieve 94% test accuracy Fraud Detection in Ad Clicks for Mobile Apps Python, Gradient Boosting

• Designed XGBoost model to predict fraudulent click traffic from online click dataset with an accuracy of 98%

• Handled data imbalance using down sampling and boosting along with K-fold cross validation to achieve high F1 score Image Recognition Python, Convoluted Neural Network (CNN)

• Built an image recognition system using CNN with Keras TensorFlow to classify images of clothing item

• Hyper tuned parameters with GridSearchCV to achieve 97% test accuracy

Contact this candidate