December 2018 Update …..TRANSCRIPT Parts 1 and 2
Part 1
Louis DiNardo: Good afternoon everyone. This is Louis DiNardo. I say good afternoon here, good morning in Australia and good evening in Europe. Thank you for joining our call. It's the year end update of 2018. I think as you know, we do quarterly updates that we publish each 4C's as well as the half year and the full year report. It seemed appropriate to me to do a year end call, particularly since in Australia you're going into your summer. Good update before you go into your summer and then you'll come back just about the time that we do the quarterly update.
We do five calls a year, which I'm not sure if that's typical in Australia but certainly here in the US every quarter, a lot of public companies hold a conference call and usually it's with analysts that ask a bunch of questions. As you know, I take questions by email. I have them all in front of me and I try to incorporate them into the presentation. Those that I don't catch during the presentation, I'll do a quick review before we end the call to make sure that we don't miss anything. This shouldn't take very long. Again, it's a year end update. You know, we just closed the fourth quarter. We don't even have the numbers in on the fourth quarter yet. I'm not going to talk about the fourth quarter financials. That'll happen with the 4C at the end of January.
Moving on through the deck if I can get this to work, there we go. Then just post the standard disclaimer that you've seen on every call. You can, for some reason some of this looks like it's some typos in there but you can read this at your convenience. Agenda for today, very straightforward agenda. We're going to look at highlights through 2018. Of course we need to look at low lights. We always have to look in the mirror and make sure we understand where we are with respect to where we expect it to be. Business update on very specific things like GPI, like SM Tech, like commercial agreements as well as some of the company names that we've talked about the past. Bring you up to date and make sure that everybody has clarity on where we are. I think, most importantly, the key to development.
I think it's important to recognize that this juncture, we're now at that point where we want to fulfil Peter Van DerMade's vision of building a more and more fixed system on a chip. We'll talk about that at some level of detail in the presentation. You know, Peter's been working on this thing for 10, 12, 15 years. BrainChip's(ASX:BRN)been at it since the test chip, the Snap 64, was demonstrated in late 2016. The culmination of Snap 64 technology, the acquisition of the JAST license where Peter has, Peter and company, the research team, have done in the way of invention since then. Those three things combined are really what brings Akida to the fore.
I'll go into this a little bit more detail in a minute but I think when you think about Akida, you should think about the edge where large volume, high volume devices, whether it's IOT, medical devices, you know, a wide variety. It could be a car, the bumper in a car we consider an edge device because you're sitting out there with the transistor. There isreally no AI solution at the edge today. You could think of this and think of, maybe a little tough to say this, but you could think of Akida or BrainChip(ASX:BRN)as AI at the edge and Envidia like at the edge. That's really what our goal is, just high volume markets with the most compelling solution.
2018 highlights, you know, in many respects 2018 was a great year. We'll talk about the low lights in a moment. We have built a strong team. We have a head count of about 35 people now. 80 per cent of them are research engineering. That is research called product development or engineering. The balance is the three sales guys that we have, a couple of marketing guys that we have, and then the general administrative overhead. The year was very, very productive for an intellectual property. I think you saw a press release a couple of days ago about the, they called it an "Army of Us" provisional patent filing.
Provisional patent, basically, gives you your priority date for all of your inventions. You put all of your claims in one omnibus provisional. Then you have a year to determine how many patents does that spawn. You could have a parent, you could have the children, you could have grandchildren. During that year, you can also add more claims to that patent. It was very important for us to get the provisional filing. Those inventions are very valuable. The value of this company really is our intellectual property not just in the spiking neuro network domain but in the artificial intelligence domain more generally.
We've hired a consultant Shepherd Group to help us with that strategy. We have probably the best patent attorneys in the country, in the US, Stern Kessler out of Washington, DC that worked with Peter and Ale and company in drafting a provisional patent, making sure that we have broad claims but specific enough that the patents will hold. They'll walk us through and help us moving that provisional omnibus patent filing, excuse me, into individual patents over the course of the next year.
In gaming or GPI situation is a wonderful one for us. The ATS system, our advanced table system, we should be production ready with the vision system in January. It's been a lot of work. When I think back to the company acquired SpikeNet in late 2016. I think it was September of 2016. What we really got out of that group was an extremely skilled video analytics team. We would not have been able to do this gaming products, you know, gaming partners international project had we not acquired SpikeNet. We have customer relationships, they tend to be end user relationships because SpikeNet was doing custom more projects. Again it was great insight into what we should do with BrainChip Studio. It gave us a good learning ground out of how to bring a product to market. I'll touch much more detail on GPI in a moment.
In surveillance, we have a strong sales pipeline. I say that and then I'll show you on the next slide that a low light is that we haven't turned that into revenue yet. I think there's no one that has as much angst about that as I do. We do have processes and people in place, such a great opening that pipeline, trials. We got a couple dozen trials that are committed or ongoing through design and inter revenue in early 2019.
A key to IC development, it's on schedule. We expect engineering samples in the second half of 2019. There was a question about why Akida had been delayed a year. It's not been delayed a year. We put an extra quarter into the schedule, a quarter, a few months into the schedule because we were getting so much input from customers, potential customers that we wanted to make sure that we really honed the product definition. I'll talk about that in a few moments. Take an extra quarter or so to get in JAST spot, product definition reduced to practice is far more important than getting it out of quarter early. With the development system, as I think you all know, we launched late in 2018, customers can develop really an assimilated environment what they will see and instil when it comes out.
An important part of the process here as we've talked about before is the Akida ability to do a CNN to an SNN conversion. Not to get into the technical reads but most of the world right now works on convolutional neuro networks. They do that in CPU's or GPU's. It's very math intensive. It's very sequential and deterministic. You know, it's a power hug to be able to take existing CNN's, convert them to SNN's, gives us a quicker time to market rather than trying to evangelise and educate. People had to do native SNN's. We'll talk about native SNN's in a moment, as well.
In the next bullet, the IP acquisition that we did, this was an acquisition of IP that we got from Praxis University in Greece. This may have been the best money that we spent. We spent $35,000, excuse me, 35,000 euros. What we have is a fully functioning native SNN for cyber security doing deep packet inspection, looking for malicious code that's coming across IP traffic. We've had it in our hands for a couple of months now. The results are very, very good. With no malicious code, we're talking about 99.5 per cent accuracy. We're working on when you don't know that it's malicious and you just see something averment, how accurately can you filter that out? That really brings us into the cyber security space much sooner than I would have expected.
Let's go to low lights, BrainChip Studio end user sales, I'll be generous and call it weak. BrainChip Studio OAM sales, I'll be generous and call it weak. There's, again, no one has more angst about this than me. I think what we really learned in bullet number three is that the sales cycle was more extended than what we expected. We expected a sales cycle of six to nine months. It's more like a year to 15 months. I'll say at the same time, that doesn't mean that the pipeline doesn't have credibility. You know, the pipeline is robust. We've got, as I said, better than a couple dozen trials going on or actually either going on or committed to start in January. We've really not lost any of the deals along the way. You know, you have to intercept these end users. It's more like a B of C, a business to consumer sales process than it is an OAM process where it's a business to business. When you're dealing with the police department you have to get into the police department. You've got to get an internal advocate. Once the internal advocate gets on board, you know, you show them a bit of an evaluation. He goes through a trial, he or she goes through a trial. Then you have to deal with the budgeting cycle.
Well, I'm disappointed in the, I'll call it sales or revenue cash receipts. Those are actually three different numbers but disappointed in any regard. Now I'm not disappointed at all in the progress that we've made with the sales pipeline and the progress we're making around the world. You know, when we bought SpikeNet we had customers in France, predominantly Paris but customers in France. Now we're spread throughout Europe. We've got trials going on in Spain, Denmark. You name the country and we've got something going on, UK included as well as here in the United States.
Part of what we had to do as we start to better understand the end usersales environment was really kind of re-tool the sales force. That's when we brought in James Earl. James is in southern California. He's lived his life in law enforcement. He knows which door to knock on when he goes to a police station. They welcome him with open arms. He's been able to get meetings, just about everybody can name the United States that we'll be on the top of your list. Retooling the sales organisation was a big part of what went on mid 2018.
Our sales forecast process, as I'm talking specifically here about BrainChip Studio, our first half is complete. It's a bottoms up forecast. It's been drilled and grinded. It's about as confident a forecast that I think you can come up with until you really have more cycles of learning as we'll get here in the first and second quarter. There's a question about product development and the cost that are going into Studio. Fundamentally, Studio is complete. We announced a revision to Studio just recently. Working on some better facial recognition algorithms. There's not a lot of engineering resources that will be expended on this. It's really about these sales organisation starting to close some waters that are on that sales pipeline.
This is more specific to gaming partners, there will be a press release that goes out on the ASX. I think it'll probably go out tomorrow morning just to give everybody an update. I think, as you know, in January of 2018 we did sign this global licensing and development agreement. That was, again, in support of their mass table systems. As you know, I think they use smart RFID at the table, they also happen to sell table felts or view the layouts, they sell the tables. This ATS is a fully integrated system that takes smart RFID, the video analytics to correlate the smart RFID and provides dashboards to the operating people at the casinos as to what's going on at the table. It's currency security and player behaviour and the optimisation of a table being online as long as possible.
Things about the gaming industry that I never knew, if you take for example the chip tray that sits in front of the dealer, if that chip tray depletes, it takes 45 minutes or longer to get chips out to a table. They have to put in a request, they have to get licensed people to get the chips out, everybody has to do their count and their recount. The ability to do a video look at that chip tray and with some sophisticated analytics, know what the betting patterns have been at the table and put in a pre order for more chips. It would never have occurred to me but that's a very valuable part of the advanced table system.
We demonstrated with GPI at the MACUL Conference, I guess that was May. Then we did GT Conference in Las Vegas in October. Customer interest was very robust. If you know the gaming industry at all, all of the major players came to a private room, saw the table system and kind of defined some wants around their specific requirements and we're adding enhancements to the video analytics, they're adding enhancements to the smarter RFID, and then the dashboards.
From our side, the video analytic system is targeted for production release in January. That is, all that we needed to do will have been done. We're very production ready. I believe the entire ATS system will be production ready and, you know, I don't know. I can't speak to when they'll have a formal announcement about the production release of ATS. There's a lot of pre-something going on with all the major players in the gaming industry. The license fee was $500,000. It was paid in full in September. There was an engineering fee of $100,000 that's also been paid in full. The original development agreement, that's $600,000 that we've talked about many times, has been paid in full. We received all of the cash.
There has been some spoke prep where they've asked us to do other things, other engineering fees associated with those enhancements will be paid for. We've collected some of that, frankly, in the December quarter. The 600,000 was already in the bank. Then there were some enhancements that needed to be done. That probably will be ongoing for a couple of quarters. Even after production release, I'm sure there's going to be enhancements as they go through cycles of learning.
I'm certain that one of the big questions for all of you listening is the commercial agreement. The development agreement was done, as I said, in 2018. The commercial agreement was expected in November. I mentioned that in the last call. The development agreement is going back and forth between attorneys. The development agreement has nothing in it that's going to change or needs to change. GPI recently announced an agreement to be acquired by Angel, which is a PR of Japan. I'll kind of step back and just an overarching comment, I think this is a good thing for us. The combination of Angel which is a leader in manufacturing cards and the shoes that cards come from, it doesn't sound like very sophisticated technology but if you play blackjack and there is a shoe on the table, it's probably an Angel shoe and it's a smart shoe. Every card has an invisible code on it. When that card comes out of the shoe, it's data logged and the house knows every card that comes out of the deck.
Angel has been working on their own vision system for a while. It has not gotten to a point where it's production worthy. When you think about Angel, which is a private company, I'm going to do this from recollection, I don't have the press release in front of me, I think they paid 110 million dollars. It was $13.05 a share, if my memory is correct. They close in price the day the announcement came out, I believe was in the high $7 range. They paid a very significant premium to potentially acquire. Let's face it, acquisitions aren't done until they're done. There's a bunch of diligence. They've got probably an anti-trust clearance. They did a very substantial premium. A good deal of that premium is, I believe, associated with bringing the ATS system under the Angel logo.
I've met with the Angel chairman and CEO a couple of weeks ago. The chairman, as well with the chairman of GPI, separate meetings and then together meetings. Everybody is excited. I saw no hesitation in the Angel CEO other than when can we get it done? How fast can we get it done? How fast can we get it deployed? They wanted me at the meeting to provide a commitment from BrainChip(ASX:BRN)that we were on board and this acquisition, if it completes, when it completes, that we're on definite board to be their partner. All in all, while it's a little bump in the road on getting a commercial agreement signed and I think that's really because they're going through diligence and they're kind of having to stop all the trains. I think by the time we have our fourth quarter update, we'll have a bunch more clarity on what that process is like. Maybe it'll be done, maybe it won't. You know, our team is working 24 hours a day with the GPI team to get this thing released as soon as possible in 2019. All good things. I think the Angel acquisition for us is actually going to play out quite well.
Now I'm going to talk about SM Tech as I get a lot of questions on this. As you know, in September of 2018 we gave away SM Tech for $609,135 or something like that. They disputed the invoice. We pursued it at a preliminary audit. I call it preliminary, it was an audit. I call it preliminary because what we got from them was inadequate. We did not get complete information. We did not get the information that would allow us to determine when, where, and how things were invoiced through Paygoo. They've been, in my view, a bit obstructionist. I have to be a little careful because we have a non disparagement agreement clause in our agreement. In order to triangulate on this, today we're filing what's called a FOIA request. It's a Freedom of Information Act through the Lockport School District. We're going to see what Lockport has on their books and records to determine who did what to whom, when it happened and what the magnitude was. Lockport will eventually supposed to give us a clean set of books and records. I don't know how long it will take them to respond. I've never done it myself, I've never done a FOIA request. My guess is it should be relatively swift.
Once we have all the information, as much information as we get our arms around, we'll determine what our legal remedies are. You know, if SM Tech doesn't pay up, then they have no what we think is legitimate reason, then we'll pursue whatever legal matters that we need. Again, I'll update shareholders at our call. If something happens sooner, of course, we'll put a press release out. I'll update, a memo, I'll update our shareholders on our fourth quarter 2018 conference call which I think will be the last week of January.
(20min delay)