Wednesday, February 8, 2006

Interview with Tim Enwall, Tendril

Our interview this morning is with Tim Enwall, Founder and CEO of Lafayette, Colorado-based Tendril Networks. Tendril is developing systems for wireless sensor networks. We caught up with Tim to learn more about his company.

Techrockies: Tell me a little bit about Tendril, and how you fit into the wireless sensor network space?

Enwall: Tendril is a provider of system software, called a “service broker”, that sits on top of low-power wireless mesh networks -- networks that are often attached to low-power sensors OR actuators. The low-power mesh network vendors have done a great job of creating viable low-power (e.g. powered by a couple of Double-A batteries) networks that transport packets. And, the vendors have done a great job of getting these to price points of under $5 (for orders of 10,000 or more). But, they stop at the packet interface.

It's a little like, in the early days of TCP/IP, if I came to you, a programmer, and said: “I want you to develop me a client-server accounting package to run my complex business.” And, your programming interface is a TCP/IP socket. That's not workable for the majority of client-server applications (you needed a client-server system software called a “database”) and it's not workable for a majority of the low-power wireless sensor & control network applications (you need system software called a “service broker”).

So, Tendril solves this problem -- ultimately speeding up deployment times, dramatically reducing the time it takes to develop an application and saving our customers a lot of software development expense. The vendors who make the reliable, low-power mesh networks -- vendors like Ember, Freescale, Renesas, Chipcon, Crossbow and Dust are either partners or potential partners of ours. We'll take the excellent work they've done making networks work and help customers make APPLICATIONS that work.

Techrockies: I see you are backed by Access Ventures and Appian--when did you last receive funding, and how long have you been around?

Enwall: We were formed 17 months ago (September 2004) -- that's when we put initial cash in the bank. We were funded by angels and venture backers. Our round formally closed March 31st, 2005 but we had a late-comer, In-Q-Tel, close after March 31st due to negotiation and contractual extensions.

Techrockies: Where did the idea for Tendril come from?

Enwall: The initial spark of the idea came from my background at Apple and by drawing an analog to personal computers. In the mid-70s the concept of “one processor for each person on the planet” was forming -- at places like Apple, Apollo, Radio Shack, Commodore and elsewhere. That processing engine had a layer of system software -- today known as an operating system -- that enabled a huge variety of different, vertical-specific applications to be built. Games, spreadsheets, accounting packages, graphic design -- a huge host of applications.

Today, we have hundreds of “microcontrollers” that are, essentially, invisible all around us. These microcontrollers are largely NOT networked today. With the advent of cheap radios and, finally, low-power networking algorithms to move packets around, we can finally begin to network the $2 microcontroller in a massive way. That “processing engine” (the network of hundreds of microcontrollers) needs a layer of system software so that myriad applications can be built on top of that core infrastructure. That's the Tendril concept.

From there, our two other co-founders -- Matt O'Kelley and Randy Willig -- both augmented the concept because of the work they had each individually done around cellular networks, Bluetooth and, more recently, ZigBee.

Techrockies: What's your own background, and why did you start Tendril?

Enwall: My background started at Apple Computer in the mid-80s after getting my Electrical Engineering, Computer Science degree from Cal. I spent almost a decade at Apple, working mostly with 3rd party developers who built applications on top of our system software. I came home to Boulder in 1994 (I was born and raised here) and went to work at a couple of Colorado software startups (InfoNow and Requisite), with a stint as the Director of Electronic Commerce at Sun Microsystem's largest wholesale partner, Access Graphics, in 1994 to 1997.

In 1998 I started Solista, a technology management consulting firm and quickly brought in a few partners to help build the business. We built the business to $5M and sold it to Gartner in 2000. During that time, one of our partners, Joel Hassell, and an associate, Steve Reynolds, started Intellocity, a Denver-based system software company focused on the interactive television market. That company grew to $7M in revenue and was sold to what is now OpenTV - the world's largest pure-play maker of system software for the interactive television market.

I stayed at Gartner for 2 years, studying the macro trends in the market -- among them wireless and the move to “physical computing” -- like RFID and wireless sensor networks. We wanted to invest some of our money in these trends, so we found a local RFID company, SkyeTek, and invested in it.

Ultimately, we started Tendril because we see a huge, disruptive transition beginning in the market -- as big as the disruption of “one processor per person” in the mid-70s. It's now the process of networking the billions of microcontrollers that ship every year. When we can network these for a few dollars each, then we can begin integrating the physical world into our computing environment like we've never done before -- networking sensory data, networking control functions and devices (switches, valves, pumps, etc...).

Because we had experienced some startup success before, we wanted to shoot for a bigger opportunity, part of one of the biggest technology waves we'll see in the next 20 years. And, ultimately, as a Colorado kid I wanted to build a GREAT Colorado company -- in the tradition of Storage Tek, Exabyte, TCI, and JD Edwards.

Techrockies: Thanks!