-
Experience
Cambridge Mobile Telematics - Software Engineer Intern
• Built the Jenkins CI system for Android Environment and integrated it with Bitbucket.
• Collaborated with various teams and gathered requirements for the desired workflow.
• Automated build configurations for continuous integration, continuous deployment, and continuous testing.
• Improved the build time by 75% by running parallel builds on AWS.
-
Projects
Mini Dynamo (Java, Android Studio, Sockets)
• Developed and implemented Amazon Dynamo styled key-value storage with data partitioning, quorum-based replication, and failure handling.
• Guarantees simultaneous availability, linearizability, and eventual consistency.
Distributed Hash Table Prototype (Java, Android Studio, SQLite) GitHub Repo
• Devised a peer-to-peer distributed hash table based storage application based on the Chord protocol.
• The application provides ID space partitioning and re-partitioning. Handles node joins and ring-based routing for insert, delete, and query operations.
Group Messenger (Java, Android Studio, Sockets)
• Created a distributed group messenger that can send messages to multiple devices using Java Socket API and store them in permanent key-value storage with Total and FIFO ordering guarantees.
Pint OS (C, Linux)
• Implemented kernel-level functions such as Priority Scheduling with priority donation, Thread yields are handled based on the priority of a thread.
• Designed and implemented a kernel-level Multi-level Feedback Queue for Pint OS.
Movie Genre Prediction (Python, Spark, Hadoop, Kaggle)
• Integrated Term Frequency-Inverse Document Frequency (TF-IDF) based feature engineering technique on a logistic regression model to predict the genre of a movie from the movie summary.
• The accuracy was further improved to 100% by the Word2Vec feature engineering method.