Building Ads Targeting Infrastructure

• Extracted insights from data attained by Archived Data Management System project (funded by LA County Metro).

• Utilized Spark Streaming to analyze incoming data streams from Freeway Traffic Sensors and persisted the results to MongoDB.

• Configured a network of Kafka Brokers and employed Kafka Consumer to retrieve data stream from 1400 Traffic Sensors per sec.

• Deployed a Web Service which delivers Near-real-time Movie Recommendations on an interactive dashboard.

• Developed a REST API (Spring) and a scalable Front-end (AngularJS) to record User Ratings and employed MySQL to store the data.

• Scheduled Spark jobs through Spring Boot application in order to refresh recommendations every 20 minutes.

• Utilized Spark MLlib to implement Collaborative Filtering and reduced the Spark Job execution time from 600 sec to 100 sec.

• Adopted Youtube Analytics to evaluate the Audience Retention characteristics and utilized D3.js to visualize the data.

• Tasks involved rewriting of the front-end of the project from AngularJS to Angular and refactoring the REST API which supported the functioning of the front-end.

• Course Producer for CSCI 360 (Introduction to Artificial Intelligence)

• Developed the Project Rubrics and Orchestrated solution for various projects and assignments which include Number Recognition using Naive Bayes Classifier, Implementing Astar search and Markov Model to obtain optimal path in a simulated maze.