Computer vision is an immensely fertile area for entrepreneurs and investors
/
Matt Turck, Partner at FirstMark Capital
Matt Turck is a Partner of FirstMark Capital. He invests across a broad range of early-stage enterprise and consumer startups. Prior to FirstMark, he was a Managing Director at Bloomberg Ventures, the investment and incubation arm of Bloomberg LP, which he helped start. Previously, Matt was the co-founder of TripleHop Technologies, a venture-backed enterprise search software startup that was acquired by Oracle.
Matt was sharing his knowledge on trends and investment opportunities in visual technologies as a panelist and a Startup Competition judge at the 6th Annual LDV Vision Summit.
Evan Nisselson, General Partner at LDV Capital, asked Matt some questions about his experience investing in visual tech.
Evan: Your entrepreneurial experience as co-founder of TripleHop which was acquired by Oracle, then Managing Director of Bloomberg Ventures, and now managing director at FirstMark Capital. Which aspects of your expertise do you believe help you empower entrepreneurs to succeed and why?
Matt: It’s pretty important to know yourself as an investor, and know which areas you can have a differentiated point of view and have something meaningful to bring to the table beyond the money.
For me it’s pretty simple…I’ve spent most of my career in software, data, ML/AI and infrastructure, so for the most part I try to stick to those areas as an investor, with a little bit of more experimental frontier tech here and there. Thankfully some of the most exciting things happening in tech right now are occurring in those domains.
By focusing on a few domains, you get a strong compounding effect over time where you accumulate a lot of relevant knowledge and build deep networks. You see the movie a bunch of times, and that gives you a lot of context.
It’s also been very helpful to have been a founder and entrepreneur. It’s harder to build empathy, and truly internalize how hard this stuff is, unless you have experienced it first-hand yourself. Having been in the trenches helps you as an investor knowing when to be active and involved, and when to back off and just listen.
Finally, beyond what I can bring to the table as an individual, the entire FirstMark team has done a tremendous amount of work over the years to build our platform, which you can think of as post-investment support. We work hard to connect portfolio companies to experts, customers and talent. It’s something we’re very proud of, which has delivered very clear benefits to the companies in the FirstMark family.
Evan: FirstMark has invested in many visual technology businesses including Pinterest, which recently had a successful IPO. Please give a couple examples of others you have invested in and how they are uniquely creating and/or analyzing visual data.
Matt: Yes, that is correct about Pinterest, and indeed the right way to think about their business is “visual search” rather than “social networking”. Computer vision is at the core of what they do.
In terms of other examples, HyperScience is a really interesting one, not just because they’re a fast-growing company, but also because they have a less obvious use case: they use computer vision and image recognition technologies to automate back office functions for Fortune 1000 companies and the government. They can automatically extract data from massive volumes of back office documents like forms and invoices, with very impressive levels of accuracy.
Another example would be Optimus Ride, a Boston-based self-driving vehicle technology company with roots from MIT. They heavily leverage computer vision, that’s pretty much of the core of their technology.
“Big Data & AI” landscape by Matt & FirstMark
Evan: In the next 10 years - which business sectors will be the most disrupted by computer vision and why?
Matt: The two obvious candidates are medical imaging and transportation (cars, trucks). The latter will take a few more years, but will ultimately happen. The former is already under way, even though it has tiny industry penetration as of today. By the way, I do firmly believe that computer vision will “enhance”, rather than “disrupt” areas like radiology or pathology, as it will enable doctors to focus on the more tricky cases.
But there are so many other use cases beyond those: security, retail (self-service check-out), industrial (quality control), agriculture, advertising, sports, etc. Computer vision is an immensely fertile area for entrepreneurs and investors!
Evan: We are both passionate about building communities. Our LDV Capital fund organizes Vision events and monthly LDV Community dinners which include thousands of members. You organize two monthly events, Data Driven NYC and Hardwired NYC. Why do you organize these events and what has surprised you the most from organizes these gatherings?
Matt: Yes, I’ve been running Data Driven NYC and Hardwired NYC for a number of years now, and those communities have grown quite a bit -- 25,000 members in total.
Frankly, there was never really a goal when I started Data Driven NYC. I had access to a nice room and thought it’d be fun to get a few people together to geek out about data stuff. I knew I wanted to the event to be free, open and inclusive, but there wasn’t much of a plan beyond that. Things sort of took off from there.
But it’s certainly been the gift that keeps on giving. Those events have been a tremendous source of insights for me, which have made me a better investor and board member. It’s been wonderful networking for sure. For some reason, the quality of people who show up at those events has been outstanding (to an extent I was not fully expecting, perhaps that was the surprise). The communities have become a great talent pool for the FirstMark portfolio.
Beyond that, there’s the immense satisfaction of doing something for the broad community, without expecting anything immediate and tangible in return, a “pay it forward” kind of thing. A bunch of people found their next career opportunity, companies were started after co-founders met at the event, etc… lots of awesome stuff all around.
Evan: LDV Capital started in 2012 with the thesis of investing in people building visual technology businesses and some said it was “cute, niche and science fiction.” How would you characterize visual technologies today and tomorrow.
Matt: Well, I’m sure whoever told you that is probably eating their words now (laugh). That time period (2010-2012) is really when deep learning exploded as a technology that is truly ready for prime time, and it works particularly well for all image-related stuff, so the timing was perfect. I think we are still early in the overall “productization” of deep learning, and there’s a long list of “take problem X and add AI to it” which haven’t been fully explored yet.
Evan: You frequently experience startup pitches. What is your one sentence advice to help entrepreneurs improve their odds for success.
Matt: The Number One thing I’m looking for in a pitch is impeccable clarity of thought. Great pitches tend to be very compelling in a “this is where the world is inevitably going” kind of way, but on top of that, the best entrepreneurs tend to be hyper-precise and sophisticated about every nuance of the problem they are addressing and the business they are building. When you ask a question, they are just miles ahead of you, they thought of everything. You can feel increasing excitement within yourself throughout the meeting that you just want to join them on that journey and work with that team for the next few years of your life.
Matt Turck (left) hosting Florian Douetteau, CEO of Dataiku (right) at Data Driven NYC
Evan: What are you most looking forward to at our 6th LDV Vision Summit?
Matt: The whole thing, pretty much. You do a wonderful job getting this community together, and the agenda is a really compelling overview of the various current and future topics relevant to computer vision.