AutoCon 2 Lookback

As Jeff Bezos, founder of Amazon.com, once said on stage about his investment strategy: 

This timeless advice is equally relevant to network engineering, where focus on the fundamentals can yield long-term success. What are the things in network automation that won’t change in the next 10 years? I believe that is a critical concept to ponder. We have so many technologies, frameworks, tools, and directions to choose from that we can’t possibly master them all. Over the holiday break, I thought about the “things that won’t change” in network automation and by extension, what we can invest our time and effort in. 

More interestingly, starting with this idea of what doesn’t change and overlaying the topics from the talks at the recently concluded AutoCon 2 in Denver, which talks fit those “evergreen” concepts well? Here are a few of the lessons: 

1. Business Outcomes Drive Lasting Technologies

It’s easy to get caught up in shiny new technologies—we get excited by the latest technology, the breakthroughs, and the new tools, myself included. But the best tools are those that solve real business problems. 

At AutoCon 2, we saw Jeremy Schulman’s talk on using AI for better Network Observability at Major League Baseball, Hari Gollapalli’s talk on driving results at PayPal, and Joe Wesch and Aaron Drydale’s journey on automating the network at QuikTrip emphasize this. The common theme with the talks is that they lead with the business problems they are trying to solve and choose the best technology after. A special mention is warranted for William Collins’s talk on ‘Building a Business Case for Automation using Informal Channels’ 

2. The Power of Compounding in Network Automation

As we all know, compounding is the process in which an asset’s earnings are reinvested to generate additional earnings over time. It allows momentum to be built on each other, and significant gains can be accumulated over time. Compounding is so powerful that Warren Buffett, at one point the richest person in the world, calls it “The 8th Wonder of the World.” 

What does compounding have to do with network automation? Early adopters of network automation have a distinct advantage. Referring back to Jeremy’s talk, they could leverage AI to produce extraordinary results because they had already automated the mundane network tasks that took time away from their daily work and the collection of the necessary network data as the foundation.

We also saw the compounding effect of lessons learned by just getting started, even if that is not where the team ended up—such as Eric Anderson’s talk on the iterations of infrastructure service delivery and Karl Newell’s talk on Internet2’s journey and the different steps and lessons learned. As illustrated by Karl and Eric, the growth might be nonlinear or the benefit might not show right away, but they started and enjoyed the benefit of compounding. 

The best time to start was yesterday, the next best time to start your network automation journey is today. 

3. Start Small and Embrace the Messiness.

Network automation is rarely straightforward. Most AutoCon 2 speakers acknowledged starting with imperfect, messy environments. In almost all the talks, the speakers start with the messy, unpredictable, and sometimes depressing reality. We as humans crave control and predictability, but the reality is often anything but. Since almost all talks have this messiness built into them to a certain degree, I won’t mention any talk in particular. But I do want to point out two talks that can offer a good solution to dealing with the messiness. 

In many of the talks, the speaker talks about starting small. In ‘Network Automation in Baby Steps’ and ‘Step 0: Test the Network!” both Danny and Joseph offer concrete, detailed steps in getting started with nonintrusive, effective ways of getting started with network automation. To me, that is a great way to build momentum and get going. 

My list only comprises a small set of the overall high-quality and practical talks at AutoCon 2. I find myself constantly going back to previous AutoCon talks, especially regarding trending and historical data. I would recommend every network automation enthusiast to at least spend some time going through the list of talks and pick the ones that look interesting to you. I promise you won’t be disappointed.