Zach is building a Machine Learning platform company upon which 12,000 developers are building apps.
This is a terrific conversation that spans how to build developer networks for a PaaS company, and numerous related issues.
Sramana Mitra: Let’s start by introducing our audience to yourself as well as Edge Impulse.
Zach Shelby: I am the co-founder and CEO of Edge Impulse. We are the company that helps to democratize the use of machine learning on real edge computing and real industrial types of systems. That is an area that machine learning is new to. There are a lot of differences in how we have to enable all these developers and companies that want to make use of it.
Sramana Mitra: Does that mean that you have a cool engine or platform that you would then have other developers build on top off?
Zach Shelby: A really common trend in the industry right now is to use developers as a way to get to market. That is something that we are passionate about. Especially when you are dealing with new technologies and bringing technology into markets, it’s more the developer who is deciding what new technology to try out.
The whole cloud movement is very much driven by developers. APIs, new libraries, and new machine learning techniques have been driven by developers.
Sramana Mitra: I have a body of writing on the topic of the Platform-as-a-Service and developer networks in that context as well. I am very excited to learn what it is you are doing. Let’s start there. What is the scope of your developer network? How large is the developer network?
Zach Shelby: We provide normal developers to make use of machine learning to solve real-time sensor, audio, and computer vision problems. What that means is we have an online SaaS platform with all the development tools that you need to be able to do that. It is an end-to-end tool that includes everything from data collection, data set management, and all the algorithms you need to solve these machine learning problems out of the box.
There is also lots of visualization which visualizes what is happening in what works and doesn’t. Finally, there’s a lot of testing and DevOps to get these algorithms deployed on real hardware. We do all that as a cloud platform aimed at the developer. That is where the developer comes in. They are our core users.
Developers use this to integrate machine learning in what they are working on in their day job. We launched this early in 2020 and as of today, we have about 12,000 developers on our platform. They come from thousands of enterprises where they work. They have lots of projects, data sets, and interesting solutions that they create solutions for.