Pooja K. Email & Phone Number

Kirkland, Wa, Us

Seattle, Washington, United States

Greater Seattle Area

• Architected a data pipeline to detect duplicate Stack Overflow posts using Locality Sensitive Hashing (LSH) algorithm• Pre-processed 50GB of data, generated the hashes for LSH on spark cluster housed on AWS and stored them in PostgreSQL• Used Postgres’ indexed arrays to search user query and Flask to host the web application

Scottsdale, Arizona

𝗖𝗼𝘂𝗿𝘀𝗲 𝗖𝗹𝘂𝘀𝘁𝗲𝗿𝗶𝗻𝗴:• Created a machine learning pipeline to cluster courses throughout the semester (before term start and after term end) in order to analyze instructors’ design effort on Canvas Learning Management Tool.• Performed data cleaning and pre-processing using RedShift and Python • Performed feature engineering to identify best features for different clustering tasks• Implemented K-means, hierarchical clustering models to identify 5 design patterns over multiple iterations • Visualized distinguishing features using D3 and highcharts. 𝗘𝗱𝘅 𝗗𝗮𝘁𝗮 𝗟𝗮𝗸𝗲:• Created a data lake to store student activity data from ASU’s online courses hosted on edX platform. • Orchestrated data from a variety of sources using AWS Glue and PySpark scripts. • Analyzed edX student activity data to better understand the behavior of students and course engagement. • Predicted student clusters and personalized emails/announcements which resulted in 3% increased retention.• Performed reporting and querying to serve researchers’ ad-hoc requests using SQL scripts on Oracle, RedShift, custom python scripts, Tableau and D3 JS

Phoenix, Arizona Area

𝗖𝗵𝗮𝘁𝗯𝗼𝘁 𝗔𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻:• Implemented a prototype myBarista Chatbot application which allows the user to place an order through voice commands and text messaging. Implemented domain specific Named Entity Recognition and Intent Classification Models using spaCy (Natural Language Processing Toolkit).𝗪𝗲𝗯 𝗔𝗽𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻:• Designed and developed full stack application for forecasting sales and marketing targets by integrating backend REST services and by using NodeJS as backend and Angular JS as frontend.• Developed components like Create/View Forecast, Create/View Promotion, Create User and User-Login.• Optimized load time by 70% by shifting the query processing to the backend.• Fixing around 10 bugs and issues in existing web applications.

Hyderabad Area, India

• Developed and designed end-to-end BI Web applications for Microsoft which includes all phases of SDLC. • Developed backend for 25+ Business Intelligence Dashboards. Created ETLs to populate over 7 databases for 3 projects. Created over 70+ KPIs (Key Performance Indicators) in SQL and MDX as per interest of business.• Administered SQL Server, optimized SQL queries by 40%. Reduced the refresh time from 9hrs to 2hrs.• Developed applications to pull data using SOAP and REST APIs and Selenium to capture page load time.• Involved in discussing requirement specifications, sprint planning, production deployments and project progress with the Clients as well as On-shore development team. Been a POC (point of contact) for task allocation, training and task completion for 4 interns and 3 graduates.

Master'S Degree, Computer Science, 3.89

𝗔𝗰𝘁𝗶𝘃𝗶𝘁𝗶𝗲𝘀 𝗮𝗻𝗱 𝗦𝗼𝗰𝗶𝗲𝘁𝗶𝗲𝘀: • Member of Software Developers Association - Soda ASU • Member of Women in Computer Science - WiCS 𝗖𝗼𝘂𝗿𝘀𝗲𝘀: •.

Bachelor'S Degree, Computer Science, 8.5/10

𝗔𝗰𝘁𝗶𝘃𝗶𝘁𝗶𝗲𝘀 𝗮𝗻𝗱 𝘀𝗼𝗰𝗶𝗲𝘁𝗶𝗲𝘀: • Organized Technical events like Paper Presentation and Code Cloning events at Infinity 2k13, a national level.