Yash Kothari Email & Phone Number

Boston, Massachusetts, United States

Boston, Massachusetts, United States

New York, United States

Seattle, Washington, United States

Bellevue, Washington, United States

Supply Change Optimization TechnologiesMy work at Amazon involved the development of AWS CDK-based infrastructure, creating CDK packages and CI/CD pipelines using monocdk and cdk-lib, which greatly improved infrastructure provisioning and deployment efficiency.* Led the development of AWS CDK-based infrastructure, resulting in improved deployment times and reduced developer work* Created CDK packages and pipelines using monocdk and cdk-lib, improving infrastructure provisioning and deployment efficiency* Used AWS-cdk to create and run EMR clusters, boot EC2 images, and store data on S3 improving infrastructure provisioning* Administered CloudFormation stacks, automating key tasks for ML projects, and improved deployment efficiency

Mountain View, California, United States

Google Maps (GEO)As a Software Engineering Intern at Google in the GEO division, I developed a new Place of Interest feature for Google Maps Mobile (GMM), providing users with insights based on their current location.* Developed a new Place of Interest feature for Google Maps Mobile(GMM), helping users get insights on their current location* Utilized Borg(Kubernetes-like) to retrieve and verify data from GWS, using Java and Android development, catering to 200M DAU* Tested and debugged features extensively to ensure stability and high performance across diverse devices and networks* Engineered a multitenant user interface on aGMM, featuring card layout, increasing user engagement and satisfaction

Boston, Massachusetts, United States

My project at Squak.ai aims at creating an AutoML pipeline that can process any structured dataset from any domain with no/minimal human intervention. The pipeline analyzes and pre-processes the dataset before performing feature selection. Then multiple models are trained on the transformed dataset in parallel. Finally, we create a leaderboard with top models along with their performance. A critical part of this pipeline is model interpretability. For each model, we compute the significant features along with their importance. Below are my contributions to the AutoML project:* Augmented efficiency, executable size, and readability of previously poorly maintained code modules through code refactoring, optimization, and documentation, covering best practices in OOPs and following the PEP-8 style guide for python programming which helped decrease the run-time of the code by 20%Skills: Python, Pandas, Numpy, Javascript, Typescript, Nest.js, Rest, React, Atom, GitHub* Optimized AutoML and Time-Series Analysis code to help ML-focused organizations transform data into quantifiable decisions* Engineered a novel way to detect and handle outliers in time-series datasets improving the accuracy of the training models by 20%* Implemented connectors for Snowflake, Google Sheets, Google BigQuery, and SQL Server as microservices and pushed data from AWS S3 to these databases* Implemented CI/CD pipeline for auto-scaling of AWS resources through bash scripts* Adapted professional software engineering practices like code reviews, source control management, build processes, and testing

Mumbai, Maharashtra, India

As a Python Developer Intern at Quidich, I had the opportunity to work on cutting-edge technology and develop skills in Python, C++, and computer vision.* Analyzed implementation of Visual-SLAM and its application for real-time tracking and mapping utilizing drone footage* Developed a feature to perform image registration on drone footage to help reuse saved spatial data of familiar sites* Slashed 2 hours of system calibration time by reinstating drones in familiar sites by reusing map checkpoints* Implemented and analyzed multiple deep-learning techniques to organize a more client-friendly film catalog of videos footage shot by the organization, thereby accelerating client retention and satisfactionSkills: Python, C++, ORB-SLAM, Visual SLAM, Pandas, Numpy, OpenCV, Image stitching, Geometry, Computer Vision, Feature Matching

Master'S Degree, Computer Science, 4/4

Courses: Algorithms, DBMS, Web Dev(MERN Stack), Program Design Paradigm, Foundations of Software Engineering, Information Retrieval, Data.

Bachelor Of Engineering, Electrical, Electronics And Communications Engineering, 8.84/10

Courses: Structured Programming Approach(C), Object-Oriented Design (Java), Data-Structures, ML and AI, Cloud computing