Why we invested in Pryon
The better part of 2017 was spent engaging with entrepreneurs working on meaningful projects. This meant meeting lots of folks, trying out plenty of beta products, helping out current Attic Ventures portfolio companies when possible, and cooking up some of our own in house experiments.
It is through that process that I met Igor Jablokov and I am thrilled to announce that his new startup, Pryon, is the latest investment in the Attic Ventures portfolio.
The company is bicoastal, with engineering and research tentacles in New York and Seattle. They are building an AI assistant for enterprises — or the true “Alexa for business”. Pryon was founded by Igor Jablokov and his previous startup, Yap, was Amazon’s first AI acquisition to help them develop Alexa and the Echo product line.
Pryon is defining what it means to be an AI enterprise company. The canvas is virtually blank, and they have put together a first in class team to serve this expanding market sector. The Fortune 500 is adapting to be reticent in working directly with Amazon given it’s continued expansion into their businesses. In fact, several of them banded together to invest in their futures instead of leaving it to the tech industry’s whims. Pryon is one of those bets they’ve made, given their realization that can’t survive or thrive without an insider’s view of where AI is going.
The Team:
Igor is a solo founder working on his second startup in a sector he not only understands, but also helped birth. Beyond Alexa, his team worked on IBM’s Watson (aka the Jeopardy playing supercomputer that beat Ken Jennings), worked with Apple on voice pre-Siri, and led the multimodal research team at IBM where they invented the world’s first speech enabled web browser.
The Pryon team is comprised of talent who worked together before, developed successful products in a nascent, yet hopeful industry, and have the domain expertise to build the next generation AI for enterprise customers — how many startup teams out there can say that?
Persist and Adapt:
Over the past months I have watched Igor present to anyone and everyone who is at all interested in what Pryon is up to, take in the learning from these meetings, and synthesizing it into his next presentation. He turns over every rock looking for an advantage and nothing is above or below him.
Big Idea:
In order for a user interface to “just work” in the enterprise, sophisticated middleware is required. Pryon will build that layer and manage the interactions necessary for more serious use cases.
Signal:
While the debate rages on about which multimodal form factor will become king sales of consumer smart speakers just keep climbing.
One in six US adults already owns a smart speaker, outpacing both smartphone and tablet adoption.
Connecting the Dots:
Back in 2011, when I first heard about Siri, it was easy to imagine a future where we all talked to our devices, oh, and yeah, I saw the movie Her. However, when I tried to use Siri I immediately returned to pecking at my phone and typing on my laptop, Siri was not ready for prime time.
What I did notice was my young kids weren’t as quick to give up on Siri. They were talking to her every so often, getting her to look things up for them, even having her tell them jokes. Fast forward to 2015 — my 11 year old was asking if he could use his own money to buy an Amazon Echo. That summer, and into the fall, I watched him, no listened to him, struggle to get Alexa to do much of anything. The one thing she could do was play music. Amazingly, that one application, play music hands free, was useful enough for my entire family to start using Alexa on an almost daily basis. Felt like a killer app, albeit a very specific one, was born, at least for us.
The overall user experience of voice, and multimodal devices, is still clunky at best, throw-the-device-out-the-window at worst, but it is getting just a tiny bit better every day.
Yes, voice is not the best interaction for all user experiences, but it is for some. Furthermore, it is obvious that some form of multimodal computing is the future for both the home and work. No matter what form factor ultimately prevails, it is painfully obvious that an incredible amount data intelligence will need to be tied together to power this future.
But won’t big tech rule AI and squash any and all startups in their infancy? Sure, that is a fair point, but Jeff Bezos shared his thoughts on that topic this past summer:
Bezos knows there won’t be one AI to rule them all, and he also knows that an optimal network of conversational AIs will be able to automatically distribute across AI agents. Unfortunately, we’re really nowhere even close to seeing this happen.
Current implementations of AI assistants such as Alexa, Google Assistant, and Siri, do not have the level of accuracy and security necessary for widespread enterprise deployments. Furthermore, they are siloed as their sponsors seek to keep end users within their walled gardens.
If we are nowhere close to making this optimal network of AIs happen who is going to build it? Sounds like a nice big problem to solve, and that is where Pryon, Igor, and his team come into play.
It has been a fun getting to know Igor and I am grateful that he has given AV an opportunity to invest in his company. Personally, I have already learned a great deal from being involved on a small level with Pryon, and I look forward to the opportunity to keep learning as this journey continues.
