I am always interested in sharpening my skill-set by taking on interesting new projects and challenges. I am currently involved in developing with open technologies in AI and Cloud.
Mountain View, California, United States
San Jose, CA
Currently working in the Kubeflow open source ecosystem with focus on machine learning model deployment through KServe (formerly KFServing) which is an LF AI & Data Foundation project. Primary focus is on the ModelMesh project which is KServe’s multi-model serving back-end, enabling high-density and high-scale model deployments on Kubernetes.
Active member of the KServe working group where I drive and review contributions, while also acting as a liaison for the Kubeflow community.
Facilitate KServe integration with Kubeflow releases, and ensure Kubeflow can deploy on IBM Cloud Kubernetes and OpenShift.
Develop CI/CD pipelines using GitHub Actions and Tekton to handle functional builds, test, and releases.
Interface with internal consumers of Kubeflow to derive project requirements and roadmaps.
Commonly give talks/presentations at conferences and meetups on a variety of topics surrounding Kubeflow and KServe.
San Jose, CA
■ Actively contributed to TensorFlow open source ecosystem projects, collaborating with the community in areas such as the TensorFlow debugger tool and the WebAssembly backend for TensorFlow.js.
■ Created numerous proof-of-concept projects and starter kits to enable developers to use both IBM and open source technologies including TensorFlow, Cloud Foundry, Node-RED, React, and Watson AI Services. These projects are leveraged in external presentations and customer engagements to showcase IBM and open source capabilities.
■ Organized and instructed several classes, both internally and externally, about AI concepts and technologies.
Led a month-long IBM Technical Leadership Exchange Micro-Challenge for deep learning where 400+ participants were guided through deep learning concepts and exercises.
Conducted several introductory AI workshops at Cognitive Builder Faires, conferences, and hackathons.
Constructed a series of tutorials and videos about AI in Node.js for the IBM Developer website and YouTube channel.
Mentored several new hires in developing open source TensorFlow.js apps for a jumpstart program.
■ Collaborated with a Stanford research team to help build and deploy TensorFlow image classification models to help combat the parasitic disease schistosomiasis.
Developed and deployed a progressive web app in TensorFlow.js and React where users can perform offline image classification of parasites and snails on any modern mobile device.
Co-authored a paper about this work in the “Frontiers in Public Health” journal (https://www.frontiersin.org/articles/10.3389/fpubh.2021.642895/full)
2018 — 2020
San Jose, California, United States
While performing my roles as an open source developer, I often contributed internally on a variety of web development projects.
■ Implemented redesign of external facing PHP-based website for Call for Code, a global coding competition that IBM hosts which attracts thousands of submissions yearly.
■ Built a custom analytics dashboard for tracking team activity and open source contributions. Angular, Elasticsearch, Kibana, and Python were used.
■ Designed and developed a custom judging platform for ranking submissions for the global Call for Code competition. This platform was PHP and MySQL based and deployed on Kubernetes. It has been used by several high-profile judges for determining the winners of the hundreds of thousands of dollars in prizes.
San Jose, CA
■ Core contributor of OpenStack’s RefStack project, the official tool for OpenStack interoperability validation.
Created the Python and AngularJS-based website for vendors to submit and manage interop test results.
Built Python CLI test harness for executing API tests against OpenStack-based clouds.
Developed and maintained Puppet manifests for continuous deployment of the RefStack website on OpenStack infrastructure.
Actively participated in Gerrit code reviews and community meetings.
Assisted in ensuring IBM products passed OpenStack interoperability requirements.
■ Worked on a performance team to complete numerous cloud related performance evaluation and improvement projects.
Assisted in turning around the near-failure of a cloud based application deployment service by identifying and resolving numerous performance impacting issues related to database schema and Java code logic, improving API response times by up to 25 times.
Worked closely with IBM Public Cloud development and operation teams to ensure the successful launch of an OpenStack based Container Service. Provided a data-backed list of tuning and recommendation of hardware specs to support the expected capacity and performance during initial public launch.
Education
2009 — 2013
University of Delaware
Bachelor of Science (BS)
2009 — 2013