I am often asked how I got the opportunity to join the Machine Learning Hero program, so I wanted to share my story!
It started with me attending the 2019 AWS Summit in Santa Clara, CA. I participated in a workshop on AWS DeepRacer, an entry-level tool to help developers get into ML. I trained a computer model for the self-driving toy car and won the race on the physical track on the expo floor. The next thing I knew, I was on camera with an announcer, interviewed for a blog post, and won an all-expenses paid trip to AWS re:Invent in Las Vegas to compete with 63 other racers from around the world (I came in 63rd of 64 :-).
Some months later, I received an RFP for a project that could benefit from computer vision, and being short on capital; I asked the AWS DeepLens program manager if they would be willing to comp us with a couple of devices to build a POC.
They obliged in exchange for a blog article, which was the start of our Deep Learning Traffic Control project. That got me connected to the internal AI/ML teams at AWS. Then I built and blogged about The Poopinator, and that caught the attention of the AI marketing folks. Ironically, they paid a production company to make a video for a new series called AWS Innovators; we are S1 Ep2. AWS paid to promote me using their technology to punish pooping dogs.
AWS was creating a program called AWS Community Builders, and I was recruited into their beta group. The program is application based now, but they were testing the waters at the time, and I was a good fit. We are continually asked to write blogs about projects we are working on to highlight our area of interest. I was asked if I had any ideas, and I ultimately connected to another CB member, and we collaborated on a project called What’s In My Fridge.
I was also invited to participate in AWS discussions with leaders in the AWS DeepRacer Community. They wanted to Open Source the tech stack so that the car could be used for purposes other than racing. They asked us if we had any ideas, and me being the crazy idea guy, I proposed mounting a Nerf Blaster on the car. Much to my dismay, they asked me if I could build it in 3 weeks, and I foolishly said yes. They ended up delaying the launch of the Open Source project so I could finish my project, DeepBlaster, which was the most ambitious of the three community projects selected. I will also be speaking about DeepBlaster at re:Invent this year. At that time, I was selected to be an AWS ML Hero, and I attribute this to engaging AWS, promoting our R&D projects, and sharing knowledge.
It doesn’t hurt that I have a lot of ideas, and I’ve executed those ideas with our staff and interns, whom we curate into employees if everything aligns. Keeping up with all of this has been a second full-time job, and one of my motivations is gaining visibility and increasing lead gen; this is working as we’re starting to see larger companies showing interest in our ML capabilities.
Earlier this year I was asked if we could expand the use case for What’s in my Fridge, so I proposed a beverage inventory system. I again received support from another Community Builder. I could synthetically generate images of beverage cans and train an ML model to classify the flavors of the soda water in the Fridge at work. AWS hired another production company to make a video series on AWS Rekognition, which resulted in a promo and demo interview with the Rekognition program’s sponsor.
Authored by:
Chris Miller, Launch Brigade CEO