Designed and implemented a cross-platform, multi-modal dialog system. Collaborated with Unity Engineers to develop and integrate the dialog system for educational games

Developed Android and iOS modules for cloud services, cross-platform interfaces, database/storage, machine learning, and core AI features

Tech Stack: C++, Swift, Obj-C/C++, Java, Kotlin, C#, Bazel, Clang, CMake, Gradle, JNI, RxCpp, RxJava, RxSwift, MLKit, Kaldi, Google Cloud ASR, Couchbase, Firebase, Realm, Cocoapods, Maven, gRPC

Developed and updated AI cloud services i.e: integration of machine learning features, database server, and principal system logic

Setup CI/CD server and pipelines for mobile, web, and internal applications

Implemented internal tools to help minimize developer bottlenecks and improve stakeholder knowledge

Tech Stack: C++, Python, Java, Obj-c/c++, Bazel, RxCpp, gRPC, Protobuf, PIP, Wheel, PyPI, Twine, Textmate, Azure, Jenkins, Bash, SSH

Maintained web development environment (IDE) used to write and deploy proprietary language scripts

Tech Stack: Javascript, PugJS, jQuery, HTML/CSS, Webpack, Node, gRPC, Firebase, GCP, Bazel

Implemented machine learning models to process educational videos and generate training data as code-blocks

Performed analysis on text to automate grading of written responses and generate statistical reports on student performance

Tech Stack: Python, Go, Node, WebRTC, RTSP, Spacy, gloVe, nltk, pyMC3, OpenCV, Google ASR, Numpy, Android, Firebase, VM, quagga

Applied supervised machine learning methods to predict hand poses given the orientation of joints.

Reduced computation cost by creating a revised model that minimized data and run-time needed to make inferences.