Sidhraj Solanki Last Updated on **th December ****
ac72j6@r.postjobfree.com 508-***-****
EDUCATION
UMASS BOSTON
MS IN COMPUTER SCIENCE
Dec 2018 Boston, MA
Cum. GPA: 3.78
M.L.S.U.
BSC IN COMPUTER SCIENCE
Rajasthan, India
LINKS
Github:// sidhraj
LinkedIn:// sidhraj
COURSEWORK
GRADUATE
Computer Vision
Artificial Intelligence
Object Oriented Programming
Data Backed Websites
Database Management
SKILLS
PROGRAMMING
Java • Python • R
• Shell • LATEX
• C • PHP • SQL
MACHINE LEARNING TOOLS:
XGBoost • KNN • Random Forest
• Image Processing • NLP • nltk
• Keras • Tensorflow • sklearn
FAMILIARWITH:
Java Servlet
OTHER SKILLS:
HTML • CSS • MS Excel • vim • JIRA
EXPERIENCE
Datycs Inc.
SOFTWARE ENGINEERING INTERN
June 2018 – Aug 2018 Boston, MA
• Web scraping using python libraries to extract data from web pages.
• Processing data files, extract information from pdf and other format files, and generate reports using python.
• Applied natural language processing systems like Apache cTAKES, MetaMap to get annotated semantic disorders from electronic health records.
• Created classification models using scikit-learn to identify performance of text extraction models.
• Evaluated performance of different machine learning models using python.
• Commit and push changes to bitbucket repository. Aptara Corp.
DATA TEAM
May 2013 – Jan 2017 Noida, India
• Worked on data cleaning and data pre-processing for Elsevier Customer Insights and Analytics Team.
• Identified target audience of clients by demographics, geographic, income, interests, and buying behavior using social media analysis.
• Advised and created social media strategies for clients to enhance their business.
INDUSTRY PROJECT
Topos Labs CHATBOT
September 2018 – December 2018 Boston, MA
• Created ontology based on the book ”Fellowship of The Ring”.
• Fed the whole book paragraph by paragraph to Edison(Cognitive Engine of Topos Labs).
• Created a knowledge base having popular questions and answers based on the same book.
• Created a chatbot using python Tkinter, giving option to users to query either knowledge base or Edison.
• Used python nltk to find nearest match for existing questions from the knowledge base.
• Wrote python script to generate elastic search queries from user questions and query Edison backend (Kibana) to return an appropriate response. ACADEMIC PROJECT
Precipitation Forecasting
• Forecasting of extreme precipitation for the state of Iowa, given historic data for 30 years. The data contained details of 9 variables for 5328 locations(North Hemisphere).
• Used SAOLA (Scalable and Accurate Online Feature Selection Algorithm) to select relevant features from approximately half million features.
• Used oversampling and undersampling techniques like SMOTE and SMOTETOMEK to deal with data imbalance.
• For prediction used KNN, SVM and XGBoost.
1