I am the lead technical contributor to Apache OpenWhisk, an event based computing model for the cloud. Since its inception in IBM Research in early 2015, I have lead in designing and implementing the core OpenWhisk architecture. You can find the source code on GitHub. OpenWhisk is the leading open source platform for research and innovation for serverless computing, which is also known as Functions-as-a-Service. It is production-ready and available on the IBM Cloud via Bluemix. It is also deployed by Adobe as I/O runtime. If you are a graduate student and interested in collaborating on research topics related to serverless computing, send me your resume.
Before OpenWhisk, I spent several years leading the Liquid Metal project, which I also helped start. It was a research effort that produced a language, compiler, and runtime for high level synthesis of reconfigurable hardware. My other contributions at IBM include computing with spreadsheets, and the content management system for IBM Researchers which is the service behind the web page you are reading right now.
Before joining IBM, I spent several years as a research scientist at MIT where I contributed to StreamIt, a language, compiler and runtime for stream programming. I later developed and co-taught a multicore programming primer as MIT OpenCourseWare.
As an undergraduate, and later in graduate school, I made several contribution and supported Trimaran, a compiler and simulator for VLIW architectures. It was particularly rewarding because Trimaran was used for teaching a few compiler courses around the world.
- Serverless HTTP handlers with OpenWhisk
- Run Swiftly: precompiled Swift actions
- Using Docker actions, running Go lang, and other fun things with OpenWhisk
- Locally debugging OpenWhisk actions without OpenWhisk
Read more of my articles on Medium.
- OpenWhisk sessions from IBM InterConnect 2017.
- Apache OpenWhisk now generally available on IBM Bluemix.
- OpenWhisk accepted as an Apache Incubator.
- Awesome OpenWhisk things.
- Bluemix OpenWhisk announced at IBM InterConnect 2016.
- OpenWhisk source code powering new Bluemix serverless computing capability released on GitHub to the open-source community under Apache License 2.0.
OpenWhisk news coverage from initial launch
- "OpenWhisk is coming into the world more fully baked [than AWS' Lambda]" – InformationWeek
- "The new open-source cloud product announced, OpenWhisk, is a shot across Amazon’s bow." – Forbes
- "IBM’s work around Swift, OpenWhisk and API Connect are worthy of attention." – Redmonk
- "OpenWhisk is IBM’s answer to Amazon Web Services’ event driven system Lambda." – Business Cloud News
- "OpenWhisk event-driven platform runs on Bluemix and has one advantage those other two platforms [Amazon Lambda and Google Cloud Function] don't". – InfoWorld
- "IBM is also launching a new cloud service called Bluemix OpenWhisk that represents an answer to Lambda event-driven computing service from public cloud market leader Amazon Web Services (AWS)." – VentureBeat
- "The bottom line is that this serverless computing technology is ushering in a new way for developers to build applications, particularly apps serving the Internet of Things market." – Network World
- "Servers? We don’t need no stinkin’ servers." – TechSpective
In early 2015, a small group of us in Research started a new effort to bring a cloud-first and distributed event-based programming service to market. In February 2016 at IBM InterConnect in Las Vegas, OpenWhisk was announced publicly for the first time with availability on Bluemix and on GitHub. This past December, platform became generally available. OpenWhisk was also admitted into the Apache Software Foundation Incubator.
OpenWhisk started as a small IBM Research endeavor, and grew significantly with two objectives: One, realize the compelling promise of serverless programming for cloud native applications. And two, seed the open source and research communities with a robust foundation on which further advances in serverless programming may be fostered. I am proud to be one of the founding members of OpenWhisk and to have continued my contributions to this work in what I think is a testament to a new and reinvented IBM. I'm also responsible for the Whisk in OpenWhisk.
In developing OpenWhisk, my colleagues and I took a vertical approach that crossed many levels of abstraction from applications, to programming model, runtime and system architecture. This holistic approach is characteristic of my past contributions to research and open source, and is made possible by working along IBMers that have both broad skills and deep expertise in many facets of computer science and engineering.
I am extremely excited by the potential OpenWhisk holds for IBM, the open source community, and academic research.