Intel

Joule, Edison, and Curie shipped before an ecosystem existed around them. No carrier boards to seat them on. No extensions to wire to them. No body of work showing what they were for.

Someone had to be able to go from the box to something running in an afternoon. None of these modules could do that yet.

I product managed the Inventor Platforms: Curie, Joule, Edison, and Movidius. I built the developer and partner ecosystem around them, and I spent a lot of the job as a technical evangelist.

Before the Joule spec was fixed I interviewed 50+ CTOs at leading IoT startups and ran five internal workshops on the out-of-box experience. The hardware and software decisions followed what those people said.

Then I made the modules reachable. Edison was certified for AWS IoT and for Azure IoT Core, each with its own documentation, developer kits, and sample code. I held the Microsoft IoT relationship through the Joule launch and certified Windows IoT Core for launch day. I organised joint partner enablement with Arm, Amazon, Microsoft, and Google, and with IBM. I wrote 80+ technical articles. Partners I brought in built 35+ Joule platform extensions. I supported 85+ workshops and hackathons across Galileo, Edison, and the Mashery API, and reached more than 10,000 developers. The carrier boards came from outside Intel. The internal roadmap was aimed at production volume.

I would start with the interviews again. Product teams heard what was hard about the silicon from the people building on it. That is cheap before a spec freeze. After the freeze, the same findings show up in sales data, and nothing can be changed.

A partner who builds an extension puts their own manufacturing and their own roadmap on your silicon. They do it when the ecosystem case is real. Recruiting them is most of the job.

I certified Edison against both AWS and Azure. A developer cares whether the module works with the cloud their company already pays for.

80+ technical articles. 35+ Joule platform extensions. 85+ workshops and hackathons. More than 10,000 developers reached worldwide. 200+ developers through the Edison go-to-market. Dozens of design wins from the customer research, back into product, engineering, sales, and marketing.

On Edison I worked with the OpenAPS team and organised the Explorer Board build team. OpenAPS is an open-source artificial pancreas. Thousands of people use it to manage their blood sugar.