The first day opened with a keynote from Prof. David Cheriton, of Stanford University. Also an entrepreneur and investor, he started by describing his investment history. He said he’s been lucky. Among his investments is Apstra, where he is chief scientist. Why Apstra? It’s helps customer the cost of running networks. “I couldn’t go to my care home without doing something about the problem of the cost of running enterprise networks,” he said. He said his criteria for investment are to:
- seek opportunities that add technical value to the world
- not be ruled by time to market
- look for a great team
- put eggs in small number of baskets and take very good care of them
He said he started working with Sun Microsystems founder Andy Bechtolsheim, and not just developed early software for him, but discovered a hardware bug. His investment history includes:
He met Google co-founder Sergey Brin when they were both rollerblading on the fourth floor at Stanford University. I helped and advised them, he said. He added they just needed the money, so I helped them raise the initial seed round. It was clear that the Web was growing enormously and needed a search engine that was better than those existing, which were compromised by being portals.
He hired Mendel Rosenblum at Stanford, while CEO-to-be Diane Greene was a windsurfing partner. The microkernel Rosenblum developed was the core technology which became the hypervisor.
Dapeng Zhu was researching network routing. Andy Bechtolsheim helped fund it. We realised quickly we had the right team but the wrong product. So we pivoted into 10 Gigabit Ethernet (10GbE) as an underlying technology for the cloud datacentre. Cisco had priced 10GbE high to protect its lower level products – we didn’t have that problem and undercut them.
The problem was that the network still a collection of boxes and cables, not a system. Apstra had lots of value because the time and technology were right, and there was a high barrier to entry.
Prof Cheriton then shared his thoughts about the state of networking. Jeff Bezos has built a walled garden, he said, like medieval computing. Is public cloud truly public, he asked? AWS is not, because it’s just a paid-for service like Disneyland. But companies use services on AWS and are then locked in – their business doesn’t work without the apps that depend on AWS. Boeing is moving off AWS because it sees Amazon as a competitor because Boeing makes drones, and Amazon is developing drones for making deliveries.
We’re in the era of a third wave of computing = pervasive automation, he said, citing examples: Uber, Waymo, Airbnb, KATERA, Caspar, Apstra…
So have I been lucky? Yes, but prepared: fortune favours the prepared mind. I focus on real-world problems. Interest in complex software systems in a distributed was, and of course networking.
- We live in transformative times
- Be on the right side of history
- Never invest in something that doesn’t add real value
- Or a startup that exists only one reason
- Or something you can’t explain to your mother
- First movers often first losers
- Be lucky and prepared, choose your friends well
Prof Cheriton was then interviewed on stage by the head of GlobalData’s research and analysis Jeremiah Caron. What were his investment attributes/criteria, and how did he differ from a VC? Prof Cheriton said VCs don’t operate like me because people want the money invested – they’re a business that wants a quick return and they can only be on a few boards, whereas I can focus on and nurture those companies that I think add real value.
Debate I – Building Unicorns: Proven Drivers of Success for VCs, Enterprise Venture Funds, Private Equity Firms and Business Angels
Chaired by Manek Dubash, Content Director, NetEvents
Panellists: Cack Wilhelm, Partner, Accomplice; Victor Chen, Associate, CapitalG; Hiro Rio
Maeda, Managing Director, Draper Nexus
Deputising for analyst Peter Burrrows who was unable to chair the debate for medical reasons, Manek Dubash opened the debate by pointing out that the real driver underpinning this debate was the equation between risk and reward. After the panel of VCs had introduced themselves, Dubash asked them about the nature and notion of unicorns, whether the number of unicorns has gone down, and whether they believed in the concept of unicorns at all.
Cack Wilhelm said she thought the unicorn term was a bit misguided, the number of $1bn being arbitrary. She said: “I think what is more interesting is companies hitting milestones like a million ARR, $10 million in revenue, $100 million in revenue, going public on a publicly listed stock exchange.”
Victor Chen said he agreed about the terminology adding that he doesn’t see a slowdown in startups. He said: “The markets are extremely fraught that you see a lot of high evaluations that come up.” He said there would be more companies creating billions of dollars of value from displacing existing infrastructure or offering net new value, citing examples such as blockchain and AI.
Hiro Rio Maeda said that the uncorn terms could be useful but only from a marketing point of view. As an example, he cited one company that was bought by HP, saying: “At that time the definition of unicorn doesn’t matter for the buyer. They just go through and scrutinise the number and try to come out with the right valuation and it came down to $600 million.”
Then, discussing a possible backlash against technology such as Facebook and Google on data privacy grounds, Chen said that the tech backlash has not impacted his company’s investment judgments. Maeda said GDPR would have a big positive impact for large corporations and startups alike. Also, he said, “machine learning and AI based application and analytics became so powerful that enterprises have been adopting them in a very accelerated way.” He said “there’s a very interesting emerging market that’s coming out of those backlash.”
Wilhelm said that apps such as Ethereum and distributed blockchains would resolve many of the privacy issues, allowing everyone appropriate access to data in a secure manner.
Dubash then asked about the geographical location of startups – was that changing? Maeda said there was something like 200 unicorns and 50% were in the US, and close to 30% in China, and about 1% or 2% from Japan. So he said it wasn’t that optimistic from a Japanese point of view. He added that in the US, Silicon Valley startups have been way too expensive. And so, many companies are starting to look at other places in the world so they can have more diverse teams. He concluded that China was going to continue to be big.
Wilhelm said she agreed about China now and tomorrow, especially its future potential because of the lack of legacy infrastructure. She added that distributed teams are becoming more and more common and socially acceptable to resolve the cost problem of hiring people in Silicon Valley.
After a question from the floor on crowdfunding, Dubash asked how the panellists would spot the losers when looking for an investment. Wilhelm said she tried to spot winners, tried to be open-minded and think that if this went well, what greatness could follow.
Maeda said he hated to see a good company that grows too fast, loses control of the organisation and becomes a loser.
Chen said it was all about whether the management team could execute. And with that the debate drew to a close.
Debate II – Can AI Solve The Internet Cybersecurity Epidemic?
Chaired by Robert Kierstead, Special Agent In Charge, Seattle Field District, US Secret Service
Panellists: Kumud Kalia, Chief Information Officer, Cylance; Slavik Markovich, Chief Executive
Officer, Demisto; Greg Martin, CEO and Co-Founder, JASK
Robert Kierstead talked about the Secret Service’s mission – who it protects, and today’s focus on cybersecurity. This helped find credit card hacker Roman Seleznev, who was responsible for $170m worth of losses. He was caught in the Maldives even though there wasn’t an extradition treaty, and was sentenced to 27 years’ imprisonment.
Greg Martin said there were not enough security people in the enterprise to protect against hackers, which is why we need AI to fill the gap – it can help them do 10 times more.
Slavik Markovitch said there was a need to automate and standardise security alerts and responses.
Following a question about how AI helps with security
Markovich said that AI doesn’t solve everything but we help security analysts be more productive. Kumud Kalia said that the aim of using AI is attack prevention not response. Greg FitzGerald said that his company automates the human tasks to help them find the bad needle in a stack of needles.
Markovich warned that hackers will start using AI to invade enterprises, and Martin said that we are in a cyber-weapon arms race, although some government energies are being used to develop these, an example of which is Stuxnet.
Conference Debate Session III—IoT: The Ultimate Driver of Digital Transformation. Or Not?
Chaired by Scott Raynovich, Principal Analyst, Futuriom
Panellists: Jeff Baher, Senior Director, Product & Technical Marketing, Dell EMC Networking & Service Provider Solutions; Michael Hallett, Co-Founder and Head of Global Sales and Business Development, NetFoundry; Russ Currie, VP, Enterprise Strategy, NetScout; Mansour Karam, CEO & Founder, Apstra; Brian Issac, Sr Director, Global Business Development, NetFoundry.
Scott Raynovich in his introduction said that there were lots of cloud-based, automated apps – even a cocktail maker. There are so many buckets of IoT tech, he said.
Russ Currie said that there is a wide range of IoT apps the main ones right now being in hospitals, and agriculture. All need securing, he said.
Jeff Baher added that all industries investing in IoT.
Brian Issac said that IoT in healthcare is complex but once you develop solutions for that, it’s applicable elsewhere.
Mansour Karam said that there are new industries and they all need to be competitive, and this is why they need IoT. He gave as one of three examples data for predictive maintenance on aircraft engines, and improvements in their design.
Baher said that IoT is growing up, that there’s a difference between IoT at work and how we use it at home and that security remains an issue.
Conference Debate Session IV—NetEvents Shark Tank Session
Introduced by Manek Dubash, Editorial Director, NetEvents
VCs: Hiro Rio Maeda, Managing Director, DraperNexus; Curtis Feeny, Partner, Silicon Valley Data Capital; Janice Roberts, Partner, Benhamou Global Ventures; Siddhant Trivedi, Investor, Omidyar Technology Ventures
The Shark Tank enables startups to pitch their company ideas to a group of venture capitalists, in the hope of winning an award in their category – and perhaps even some funding too.
The pitches came from:
Hot Startup – Cloud/DataCenter
AI Sense – WINNER
Hot Startup – IoT
POLTE – WINNER
Hot Startup – Cyber-Security
JASK – WINNER
All proceeds from the awards plus donations during the Awards Gala Dinner have been donated to three charities: Prostate Cancer Research, STEM and UNICEF.