Panel Discussion: Robert Neivert, Rashmi Gopinath, Ella Li, Xiongwei Zhou | SVIEF
12:49AM Oct 1, 2018
Now we're going to welcome our panel discussion moderator Mr. Robert neivert, followed by Rashmi Gopinath of Microsoft ventures and Ella Li Zhongguancun Investment Corporation and
Joe Zhou, he will be coming on stage to join the panel as well. Thank you again, panelists.
All right, thank you. And welcome. I know it's heading towards the end of the day. So I'll hopefully make this interesting and well worth your time. What I'd like to start is we have some great panelists here. So I'd love to let them have a little bit of time, maybe 30 seconds just to introduce yourself. Tell us a little bit about what you're investing in.
Hi everyone. I'm Rashmi Gopinath I'm a partner at m12, formerly known as Microsoft ventures where the equity investment arm for Microsoft we primarily focus on enterprise b2b investments. So Cloud infrastructure AI machine learning cybersecurity DevOps IoT business as we were pretty broad focus area, we typically look at investments in cities, HPC companies check size between two to 10, we can either leave it on a co investor and around the fund has been around since 2016, so be able to be been here for the last two and a half years, but we have been pretty active done 65 deals in the last two and half years worldwide. Our teams are based here in the Bay Area. We have presence in Europe, Israel and India.
Yeah, we're super excited to to look at some of the great companies that we have met in the last couple days here.
Hello everyone. My name is Ella Li, and I'm the Senior Investment director of zgc capital and we have a branch in our Santa Clara college zgc Capital Innovation Center at Santa Clara so
our parent company so john john data manage the junk bonds when the largest Instant TV park in China with 300 leads the firm's 40,000 patents and 30 Yoni course john Bruns who is also called a China's Silicon Valley. So under our parent company, we perform under investment and incubator models. data provided both of the incubator and management services to the top tier startups all over the world. So
investment wise that we have invested in several local funds, including down and we
also have fund of funds that are investing in seven top us funds, including XL k PCB, mellow of a piano, so forth, so the field way. So the vertical way covered include a healthcare, ai FinTech energy, and so forth. Thank you. Glad to be here.
Hi, everyone. I'm xiongwei Zhou from Boston Angel club. So I come from the far east coast. So I'm very glad to come here. The first time I attend this svief conference. So
what am I doing is a setup code of Boston Angel club is a in investment vehicle. We focus on three areas, medical device, biotech, and also a robotics. So because of Boston area have a three great advantage. One is the education for more than 100 universities in Massachusetts, more than 300 University, their second no answer. biotech, it's a world a lot of a biotech startup Innovation Center in Boston area, the sort of learn is the AI robotics center in Boston. That's why we focus on there. And after we invest early startup company, then we'll include it in Boston area. Then later on, we'll combine them is the Chinese market. Thank you.
Great, thank you. So let's get started. So what's exciting to invest in? What is it if you get an email on this particular product or subject to go? Wow, I'm going to answer that one. What gets you excited
so far as I mean, there's a bunch of areas that we consider as high priority areas, both for Microsoft as well as for m12.
These are also some of the most overhyped areas. And so we see a ton of companies claiming to be best and read best in class in these areas. And machine learning definitely tops that list. And again, because I said that these are overhyped areas, we see companies that may be analytics or some sort of pattern matching claiming to be AI. But for us as a company, it's definitely one of the highest priority areas that we want to focus on both for our internal developments as well as for investments and partnerships. The second is cybersecurity also super crowded space, but one that is incredibly important for any cloud vendor for any enterprise company. And the third, if we think about cloud migration, still being in its very first phase of evolution, given that even with a number of companies that have adopted the cloud today, the percentage is still pretty small, it's less than 5% of all workloads that run in the cloud. And so enabling cloud migration and enabling new applications that can bring a lot of those applications to the cloud is super important to us. So I would say those three are some of the key areas that we focus on. Great.
Yeah, so are there a lot of interesting areas to look at an AI is definitely one of them, it has really have a lot of impact to you in a lot of areas and also industry. So cybersecurity, definitely. So one of the ways is application VI, a, to really make this a security system much more proactive and also decrease a lot of cost of that and also health care with the API. It can also make the prediction and also diagnosis process more customer roster to a human to patients, Yost and also the transportation Ito will also make this driving process much safer. So these vital for all those reasons, like a concerns about the weather. One day, I would somehow take away a lot of a human's a job, and also increase the an employee employment rate, there is definitely a
there's a bigger place for AI to moving towards more, more more good in the world.
I hear the next panel is going to be moderated by an AI. So we'll see how that goes. How about you?
Yeah, from me, I've seen as a the advantage as a Android and went through so you can access all kind of early start off or the cool technology special in the in the Boston area, right. And since aware from the internet era, we step into the life science error right now. So I'm very exciting to see how the AI technology are changing our leaving style right now. Yeah, that's why focus on the AI robotics. So make people to live a much better and they use a
great, great, so lots of AI, and then definitely some stuff in the cloud and some definitely some stuff as well. Let's talk about what you look for in a company. So a company comes to you and they have this AI or they're bringing something to the cloud, what are some of the key things they would have to have team product market? How do you how do you pick which ones you're going to work with? So
their key factor when we're considering any deep tech companies are looking at an infrastructure company or a deep tech AI companies, obviously, the technology and the product and how real it is, and how differentiated it is from a broader horizontal AI platform perspective. There's obviously a lot of investment that goes in for Microsoft and Google and Facebook and others. And so as a startup, what is your differentiated offering, how are you bringing something to the table that is not easily available to these large internet giants that are all building their own AI capabilities. And so a lot of areas that we're focusing specifically on the AI side is companies that can bring a differentiated vertical offering and so bringing up and more defense actually having a more defensible vertical based industry specific offering that can work well with a horizontal solutions like a TensorFlow, Morrissey and TK can offer but very leveraging their industry experience. So I would say that the key for us is really the technology differentiation. The second when it comes to AI is access to data training as an incredibly important challenge that not many companies are able to overcome. And many startups are probably too early in that journey to think that they can just get customers or enterprises to share their data willingly. Data is a big mode for most companies, and not that willing to share their data with with a startup or even with another company. And so how do you have access to training data to make sure that the models that are building are going to be effective and accurate and can stand the test of time. The third is obviously the team and the background and having folks that have deep expertise in AI and deep learning would be would be helpful as well.
So sort of fit data and team
great How about you so
there are a lot of factors are we looking to before make the you management decision. So for for us, one of the three most the key elements will look at the first is a product what technology second is the management team and a third parties whether the kind of money you want to invest your face, your company, your investment companies, philosophy, so proud otherwise, what technology you want to choose a really unique type of technology is not only mean as being new and a normal but also really hard to be replicated by your competitors. And it will differentiate yourself from other competitors to not only wanted the type of technology unique. But also another differentiator is a pricing differentiator. And especially when your product doesn't really have fall into the category with a really know what technology you can have the pricing strategy, email, last word or, or more for the premium pricing and a third party is you want to find a really niche market is a great if you can find a market of that and no one has ever found, which is a you know, really rare scenario where else you want to find a marketing that a particular lay. And also perfectly face Where's your company's of philosophy, or also our product. So this is about the product and
the second part is our management team, you'll want to the team you want to have each member of the team to be really strong member. So if you'll have one member well week member, you better give some time to help improve the capacity the ability if doesn't work, then you may want to just get rid of that person, he's totally fine for this a VC company to to, to, to learn that a part of the investment will goes to train the personnel or heroin your key personnel to the company. So this is a management team. The last part of us is about a philosophy all the VC company you want to as a start up you want to first figure out a water really the UMass many companies asking for is like a financial returns water they are asking for or whether they do have asking for the strategic approach. So that part is also very important. So what are the philosophies for yours? Oh,
in fighter right? When you serve the inventor companies? I mean, where's the people? Right? So the first criteria for me is, are we looking at the founding team, it is the most important factor for us. Second is the core technology. I like the technology rather you're in this in your certain area, we will technology is a top advantage that the second the third one is the I will evaluate your product that do they have the future market in China later on. Because not only invest in the in the United States later on, because we have a lot of results in China, I will bring the China market to you. So the ability to expand to China. Exactly. That's very important to us. You know,
okay, great. Okay, let's try the opposite. What is a terrible thing someone brings it to you either something about the team or a product where you say you can't or won't, that's a bad thing. And, you know, obviously someone has a week team, that's, that's an obvious one, is there anything particular that as your investment philosophies you tend to avoid,
um, I mean, if you're coming out in a space that already has a strong incumbents are has a lot of highly competitive space, a competitive space, and you really have no differentiation in terms of tech or team our value proposition. I mean, that would be very easy.
So sort of a company says, I'm going to go after the big player, right? Straight head on. Okay. And that that too, is a tough thing to invest in.
I mean, unless you have a really differentiated strategy of how you're going to win it. If you're just going to say, I'm going to charge less than somebody else, then not that interesting.
Great, how about you? What do you think is terrible idea when it comes? So?
So I will answer the question from maybe a broader scope. So
what what a AI so a lot of people would think it was like, from, from the news coming from radio, from TV, I mean, people are thinking like a were living in a world that AI can really solve all the problems. And I would make example, in the medical field, despite of all the contribution that he has a really like a port in the resource for you in the field. But physicians or and also professionals have already realized AI is really not everything. So for instance, at this stage is still too early to say that
AI could really replace physicians, for instance, to deliver some diagnosis results without this emotion that the machine can not really replicate emotion. And also this empathy, there's, this can not only be done, and also for the for the
and also part of I would not really expand on that part. But it's also part of the reason the IBM Watson platform that was a user to recommend the diagnosis results to physicians, the reasoning, the failed is all because it cannot really gain the trust of reference from physicians. So those are particular type of the areas that when we invest, have to really pay extra attention
about you, what do you what do you avoid with a bad thing, someone brings us a bad thing, for example, right? A team come to us at all. With three of us, we have our own, we have a full time job. But we wanted to start off, right. So the party time, the part time startup badly if we raise money than we quit our jobs, and we will fully fully give us $10 million dollars and will quit our jobs. Exactly. That's the problem in the Android area is called a three I have
three I have a theory, right? The frozen family and friends and Sunday was always a fool, right? So you have a building or your you have your passion, you're you're older result star, and then you show us and said, Oh, we needed the we see. or Android or money, right? Not that that's a while. Wow.
So So part time teams attacking markets head on, not having clear differentiation. And pretty much all agreement that a lot to do with the team, I'll give an interesting one that we have at 500. And that is when people bring us an AI solution that doesn't require AI, it's, they have a piece of technology, and they're sort of hunting for a reason to use it. So a lot of times, we always say, what is the key problem. And how much advantageous is AI to this. And it's something we see a lot where they either know AI, and they're trying to find a use for it. Or they just assumed somehow, if they just play around enough with the data, they'll get a good solution somehow. And those are both warning signs to us. We very much like to see what the problem is and how big of an improvement it really is with AI. And that includes scalability right? And back to your point about data, I believe you brought up data data is one of the key things access to data is very important for us as well. We're very resistant, if you don't have access to data, let's try a little bit of a different spin. So of your AI projects you've invested in in the past, can you pick one that maybe was maybe not such a good idea, something like something to tell other investors, hey, here's a, here's something maybe you should think about for the future bad either good learning experiences you've had, or bad investments you've done and and what you learn from them.
All of our AI investments are fabulous. So of course, of course,
well, then I've got a few. But Okay.
Um, no. But I would say that one of the learnings that we've had from,
I would say the earlier stage investments that he does it, typically we start investing at the CDC stage. But there have been instances where we've come in at the seed, especially in the AI space, because we know the valuations just get all hyped. Once they get into the CDC stage, one of the learnings that we have had, from our seed stage deals that we have done is the the team may be fantastic, the tech is amazing, but the market is not ready. And so in those cases, that you are ready to market to market. And it's really about finding who are the early adopters working with them patiently till they get a product out in the market. And knowing that this is going to be different from the traditional seed stage deals where the product is the product market fit is done in nine months, and then CDs as race, but it's just going to take a little bit more time, it's going to need more help from the board level as well as educating the market and customers along the way in convincing them on why this product makes sense. And it's just going to be a longer process, right? So that was one of the learnings that we had from the
sort of the advice to them is, if you're in very early work with the customers more carefully in the beginning to make sure it's good product market fit that sound about right, how about you any you've had, I know you're perfect, and you've probably got 100% great returns. But if you could imagine a something that went wrong, what might you tell another investor about AI to say, Hey, watch out for this. So
So first of all, for the a mask and you want to my opinion to a wider the application of AI in general is exciting. So you'll do under to some specific or niche area like say, the medical imaging, self driving, those are all particular area that really need the AI the data is really important. And AI yours can really add like valuable actual value to others business. And also, we also have to realize that and all the all the AI yos can really generate a significant return on the marketing. So imagine you're both of the pieces are for the delivery type of the company and the one company just like using applying this AI to increase or decrease the cost over the repairs. Or by 10%, you may have a really like, have a great have some increased market value to a certain percent versus another company that are really using AI, but helping to decrease at the right time by like a 60% or 70% of those. So that'll work. But definitely a huge
should be a really big improvement. Basically, little improvements aren't so good for AI is not really useful. But also I think you said a specific solution, not a generalized agree mostly, we see the same thing a generalized AI is extremely difficult. And rarely we as 500, we don't have enough money to carry that to the end line. We usually we are looking for specific solutions. And that helps us fund it much more easily. How about yourself? Yeah, for the AI, right? Sometimes? And
what are the investor, right? You can solve a very simple problem, you don't need AI, right? Then when you raise the money raised some some teams at all, we are AI and so it's a look at the core, it's sexy, then it's easy to get your money. That's not right, right? What do I start where you have a lot of big data and your training the model the most important scenes, so using the AI to solve a certain problem that you're hot, who will be who will be the customer will pay you that's more important, right? Not can only evaluate AI, how the data looks, how sexy how beautiful the day that is right? Nobody will pay, you know, marketing there, right? That's the same
people shouldn't hate them, people should want to pay for the product use it because that is that's a great sign for whether a product is usable, right? Or valuable is if someone's willing to pay for it, maybe even before it's finished. So for 500, we do the same thing. We oftentimes look for early letters of intent, basically, people that are willing to pay money and how much for products delivery that helps us judges value. So okay, let's switch gears a little bit. What's what do you see AI changing in the near future. So we've done a lot of investments you've worked in the space for a little bit to something of some of maybe your own portfolio companies receive them really making impact on our lives in the next six months to five years, let's say
I would say in the near term voice is definitely becoming more and more important in every facet of application speed enterprise, the consumers and we're seeing a lot of companies that are emerging in the digital assistant for x category. One of the investments that we've made is a digital assistant for the CRM other Sales Users. And the biggest challenge that if you think about any CRM solution today, be it Salesforce or others is it's not really something that works well with how sales reps travel. And nobody nobody wants to put data into Exactly, exactly. And people typically go in at the end of the week, end of the month, but in the bare minimum information that they need to get their commissions paid out for. And it's garbage in, garbage out. And so when you're trying to do predictive analytics on the data that goes into CRM, if it's not captured at the point at which it is most relevant, you're not going to get the most valuable insights out of that data. And so voice becomes a really important tool to capture that data to put in the most valuable notes. But at the same time, that voice tool should also have the intelligence and the knowledge graph to map it to the right fields back into the CRM tools. Like you're saying something like I met this customer, my opportunity was this updated to this, it should know exactly what opportunity means in that Salesforce context. And given that most CRM tools have more than 80% of fields, customized for their own specific needs, it becomes even more paramount to understand what is important to that specific account to that specific customer. And so the,
the application of voice in AI, the application of computer vision, and AI is definitely becoming a lot more pervasive that I'm really excited about to see what what comes next
voice new CRM. So now I can yell at my CRM instead of my boss. Alright, sounds exciting. But definitely I can see the the real value especially for people that really don't and aren't around keyboards, right? They don't sit at their office all day long, they're moving. And so the
introduction of voice and as we can see in the screen behind us, it's certainly getting pretty good. All right, I'll see if I just say random words, if it can actually get it pretty good.
But how about you what's what's what do you see coming in the next few months. So
first, I, I have all the confidence for the big impact of AI in changing the cybersecurity field in America, the slick a really huge problem last year, there was almost a 1 billion separates creative rage and the government and also company are struggling for that. So he has a, we'll have a really huge influence, since it will make this ways are there some machine learning and automation capacity to will make this system much, much more proactive. So it can really detract others a virus attack the like a way ahead of the time. So that could be really a important years and other yos actually see in the precision precision medicine. So compared to the traditional Madison that are the same course of treatment are used to patients with a similar disease, state AI would combine both of the generic and also environmental data and then to predict and also make the diagnosis with human being. And also, of course, the self driving self driving field is
is a really is a really, really important ways as a application eater will make this
process much safer compared with the traditional way the driver will keep being distracted by a lot of going out looking at the radio on the put on mascara and argue with her husband wife in the vaccine. So this will be
on say cancer last night. Okay,
yeah. So So and of course in some it will always being there the potential to be used in some really dangerous feel like a mining and also trying to clearance out some radioactive materials for sure. So journalism, some medicine is seems to be like you're excited about the medicine has
I mean, it does it it has a very specific and it has a substantial advance
How about yourself? Where are you really excited? What's the What are you seeing?
Yeah, I think that is truly the cover most of them, right. So I truly believe I will try it will change our lifestyle. So read first. But I think right now how the AI can help us lead investor to identify the startup companies, right. When I was thinking about the way we can build a chat box at all before I interview you the startup you can talk to the child born with or evaluate your shoes actually enhance or not, then as I'm thinking the how the AI help
investor to identify the good startup companies that I'm seeing here.
Awesome. Great. So we're almost out of time. So we've got one last question and this is just kind of a fun one How about a long term what really is like you think wow, AI is really going to help us and this is 510 years out long term so I will personally say I'm really would love to have autonomous vehicles because honestly I'm not that great a driver so I'm looking forward to that myself How about yourself What are you really looking forward to that I will make your life better
I think going back to again the whole voice and computer vision based
technologies and I'm super excited about it's really looking at net new applications that will help us completely transform the way we go about our daily tasks like waking up in the morning and just knowing having my Alexa or Google Home interface talk to me about here's what my day looks like these are the meetings I need to be be at maybe I don't like that
but ok I'll go with it maybe you do but Alexa telling me I'm already late for my first meeting is probably not what I want to hear but
ok ok just knowing like where you need to be and what you need to plan for and just how much our daily lives are going to be changing and transforming certainly a lot where we
step away from the screen and become exactly interactive right you don't need to keep looking at a screen to see a lot of things okay so migration really from screen time to voice interacting with our computers in that way Okay how about yourself What do you see really changing your life
uh. So I
so many things but coming from a life science background I have also a particular appreciation for the bigger years were like the impact of AI in the health care field customer as Madison water precision medicine definitely Isa field that I have really the
be confidence even though at this stage is still too early or not practical to replace physicians for diagnosis wise under there's a lot of trust me to be need to be appealed about a year long Ryan can definitely is really also depends on how you're going to really combine the both of the technology and the humor you do not want to replace human beings were those physician but how to really how to really empower uses technology to empower the the physician diagnosis process and those will predict enabling them not replacing them and others sort of like the maybe the physician still heads up but the AO mentored it takes care of tasks it makes sure things are followed up on it augments
okay so augmented medicine great How about yourself
yeah for me I didn't feel AI champion my life too much till today so but I can imagine later on maybe Chandra a lot, but not today. These days. I didn't think that AI was is fully was can be used in the daily life right now. So that's my feeling right now. Great.
Yeah. So I think we're just about out of time here. So I wanted to thank you all for your time tonight. And I certainly hope everyone here got something useful from this session. But thank you so much.
Thank you. Thank you.