Gary Kalaci
on seven languages, Alexa Translations, machine translation, and CanLII headnotes
The podcast version. It also plays in any podcast app.
Transcript
6,023 words · about 26 min to readMark the passages that stay with you: press the number beside a paragraph. Marks are anonymous, and the most-marked passages help choose what goes into the book.
To hear a passage, press ▸ under its number; it plays just that paragraph.
See a misspelled name or a wrong word? Select it and press Report an error, or email the correction. The transcript is machine-made; corrections are welcome.
Hello everyone, this is another episode of the Yunusov Law YouTube show, which I now refer to as The Yunusov Question. I'm trying out different names for the show and let me know if you have any comments on this new name. We always have interesting guests. Today I have a really interesting guest that I have so many questions for. His name is Gary Kalaci. He is the CEO of Alexa Translations. And he's also a lawyer. I don't want to take any more time, and I just want to pass the floor to Gary and let him introduce himself. Hello, Gary. Thank you for being here.
Thank you for having me, Pulat. It's exciting. I've been watching your, all your previous sessions, so I'm glad that I'm on the list.
Of course you are. You are one of the people that really interest me And you know what, when I was scrolling through your LinkedIn profile, there are so many accomplishments, there are so many appointments and tenures and projects, but at the very end, something really impressed me. 7 languages? Really? Wow! So let me go through the list. I love this in people. I'm also into languages, of course, and let me go through the list. So Albanian, which is your mother tongue, right?
That's right.
Your first language. And then English. Who would have thought?
Yeah, who would have thought?
Exactly, like, right? A lawyer in Ontario. So, and then French. Comment on that? Brief comment?
So before I actually— before moving to Canada, and I was 17 when I came to Canada, I said, I'm actually, you know, my parents made that decision. I said, okay, well, I'll be moving to a country that's bilingual. I better start learning some French. So I went to the Alliance Française and started maybe about 4 months, only to realize that coming to Ontario I don't use French daily or I wouldn't be, you know, even going to high school. I was like, none of you guys really speak French with each other. I guess that's normal. And only to then go and do spend a summer in Quebec to truly do an immersion.
You know, this is so interesting. We apparently have a lot in common. I also learned French at Alliance Française. I understand you went to an Alliance Française branch in Albania though, right?
That's right, that's right.
So I went to an Alliance Française branch at Spadina and Bloor, but I also went to the French immersion program when I was in law school. So when I was at Osgoode, after my first year, I went to a very small town in Quebec called Jonquière and I spent 5 weeks. Is this the program, the Explore program, I think it was called?
That's it, I did that in 5 years, same concept.
I'm very happy that we have this program. It really helped me a lot. And then the next language is Deutsch.
Sprechen Sie Deutsch? Ja, aber
nicht sehr gut. Warum? Warum?
Exactly. The trouble is that I actually made quite a bit of progress before coming to Canada. And then for the last 20 years, I just haven't used it all that much. The trouble with languages is that, you know, the theory is you lose about 25% of what you know every year you don't use it. So you never go to zero, but 25% of what you know, it's, it's, you know, it's— as time goes on, you end up losing it. And I've been to Germany a few times since I moved to Canada, but you know, how much do you use it when you show up to a country for a few days, right? So
German might not be so popular in North America, but it's quite popular in Europe. Correct me if I'm wrong, because Germany is so important in Europe.
Exactly.
And Greek?
Yeah, so I was born right by the border with Greece in Albania, so that was— Greek was actually the first foreign language I learned. The— it's also the only language from the list of languages that has a different alphabet. So took a bit more effort, as you can imagine, because now you have to learn a different alphabet altogether.
So,
right. I enjoy Greek and I have family that lives in Greece and I go there often.
That's amazing. And the Greek alphabet, I think, is the father of the Cyrillic alphabet, which is the alphabet of my first language. I will let the audience guess what my first language is.
I
know I have an accent, they can figure it out. I gave enough clues. And then the next one is Italian.
Yeah, so Italian became almost a proximity piece, much like Greek, but also watching as a kid, watching a lot of Italian television, you know, before I was maybe in— till about grade 3, grade 4. Grade 4, I actually started studying it formally, and I did for about 10 years, even did it in university here. So when I became a court interpreter in Canada, I got certified for Albanian, but I used to do Italian as well.
Wow. So you speak Italian at a level sufficient to be a court interpreter?
At least I used to. It's been about 10 years now that I haven't been doing—
Yes.
Yes.
This is very inspiring. Very inspiring. And then finally Spanish. What's the story?
So Spanish was more for the fun and the passion. You're in North America, you're surrounded by Spanish a lot more than you were, you know, to some extent in Europe, but certainly more here. And being that it's a Latin language, Italian and French have helped with it. And I still have a lot more work to do on Spanish, let's put it that way.
Yeah, so do I. I also want to learn Spanish after my trip to Barcelona, but I actually spoke Catalan in Barcelona. I started learning Catalan. I wish it was a funny story, and I wish I could say it was— I started learning Catalan by mistake, but I intended to learn Spanish instead. No, I actually wanted to learn Catalan. So with this background, which sounds like a very language-oriented background, humanities-oriented background, you end up in University of Windsor in the Bachelor of Science program. What did you actually study there and why?
So I was in the biochemistry program. Because I was considering going into medicine. Despite enjoying languages throughout university, I did really well at sciences. My father is a pharmacist, so there was this affinity to the health sciences, you know, because I did well in them. I thought, I guess I should do— I should do medicine, you know. And it wasn't until the summer between high school and university is when I got certified as a court interpreter. So I just started getting all this court exposure, and that's when I started thinking, maybe I enjoy law more.
Aha. Is this the story of how you ended up going to law school? Because you worked as a court interpreter as a young person.
Exactly. And after second year of undergrad in biochemistry, I started my original translation company. Back then it was called Kalaci Translations.
Kalaci Translations.
So then I realized I realized that maybe I shouldn't be going to— I shouldn't be in science and medicine after all. I should do law and business because I seem to enjoy this court exposure and I seem to enjoy the business side of things.
I want to draw everyone's attention to when you started your company, which you are the CEO of and which is called Alexa Translations today. You started in August 2002, 18 years ago. Speaking of things that last, right? So you started that company when, when you were in your undergrad?
That's right, that's right.
In your chemistry program, correct?
That's right. And it was as I was working as a court interpreter, I said, I think there's an opportunity here.
So you started this company as a one-man shop, basically selling your own services, and it's nothing like that today, of course. Absolutely nothing. It's a global conglomerate today. We'll talk about that. So you went to law school after you were exposed to the courts. What did you see at the courts? Like, what made you decide to be a lawyer? What terrible things did you see there?
Well, the exposure was largely criminal and civil litigation, because as an interpreter, that's largely what you're doing. And some, some immigration cases, a lot of exposure with the Canada Border Services, immigration, because you will find both being in Windsor, you're also at the border as well. So a lot of interesting— so immigration law, criminal law, civil law, and a lot of interesting aspects. And I said, you know, I think I like— and I was also volunteering at a hospital at the time So I started sort of getting exposure to both sides of the equation. I'm like, which one do I like better? And there seemed to be more affinity to the courts, despite, to your point, not everything is all fun and like what you see on the law shows, like Harvey showing up and saying, this is why we should win this case.
And not everything is, you know, not everything is the way we experience them in the law shows that we see on TV.
You know, I also find that we have that in common because I decided to be a litigator only after I actually went to courtrooms and observed proceedings. Law school itself gave me absolutely no idea what I wanted to be as a lawyer. It's only after getting some practical experience that I decided definitely that I wanted to be a litigator. So you decided to go to law school. You went to law school in Windsor. You started in 2005. And you did a joint law and MBA program, right?
That's right.
So why did you decide to do a joint law and MBA program? Were you hedging your bets? Were you not sure about law as a career?
No, it was actually, you know, as I started the business in undergrad, I actually realized I have no formal business training. I mean, I had good math skills because being in the sciences, but running a business is not just good math skills.
Eastern Europe and math skills, tell me about it. I know that story.
So the thinking was that, you know, for an extra year to get an MBA, you also have a master's level program, because let's not forget law is still a bachelor's despite being a professional degree. And so I said, this makes a lot of sense. I get some business training in the process, and perhaps I can do corporate law because that sets you up nicely for it. I was also considering IP law because again, undergrad being in biochemistry, you're well suited to be doing IP law as well.
So you finished your law school or the joint law MBA program in 2009. By that time you've had Alexa Translations for 7 years.
That's right.
How did your company evolve over that time, by the time you finished law school?
So by 2009, I was still doing some interpretations myself. And by the way, 2009 is when we— up until then was Kalaci Translations and the switch to Alexa happened then, which I have to tell you the story as to how we ended up with Alexa Translations as well. No affiliation to Amazon whatsoever. However, we had the name first. Not that I'm positioning for the—
Tell us, tell us the story. I really want you to speak as much as you— as freely as you can and tell us all the stories. That's what we're here for.
So, Kalaci Translation make the move to Alexa Translations in 2009. I'm articling, I'm getting a lot of exposure to corporate law. Worked with some great lawyers. The, but what I realized during that process is that I'm in a situation where I'm helping other entrepreneurs build businesses. And here I have a business on the side that I'm not paying enough attention to. I mean, everyone that's gone through articling knows it's not exactly a time to run a side business. So as I got close to the bar, I had to make the decision of which way am I going? And You know, once you go to the bar, you also have the flexibility. Okay, now you're a lawyer, so you can practice as much or as little as you choose.
But this is the time to— if I'm going to give this business a real shot, this is the time to do it. So up until, you know, even through law school, etc., I wasn't doing as much interpretation or translation myself. I actually had a roster of people that I used to send for interpretations at courts, immigration, whether it was at the border or hearings otherwise, meetings with lawyers, etc. When 2010 came and I would go to the bar, I said, okay, this is it. I need to do this for real. We got our first office in January 2011. In January 2011, that was the— okay, now let's build up something that can really matter, like a real business.
And
that was also the turning point where I stopped doing any interpretation translation myself. And I said, now it's just actually building up this business. And the impetus behind it was I understood professional services well, having been in those shoes, and let's cater to the professional services as well as we can. So it was things like providing translation of documents for corporate transactions, for securities transactions, for all kinds of civil litigation cases, for IP matters, financial services, marketing services, government work. So really going after a much broader space. And to this day, we're now actually— French is by far the number one translation language, and then Spanish and Chinese. And any other languages I do or I used to do are somewhat irrelevant to our business now because they're a much smaller component of our business.
So in 2011, you made a strategic decision to grow Alexa Translations as a business, as an entity separate from you. Is this when you incorporated it also?
The incorporation was actually in 2009 because I had already made the decision that, you know, I need to grow it more and it shouldn't be just a sole proprietorship. But the growth and that's when we also thought of the name being Alexa Translations.
What's the Alexa name story? And I know that Amazon came up with it, at least published it after you.
So there's nothing, you know, people have made jokes, is that an ex-girlfriend? It's none of the above. We actually wanted something that had lexicum or lexicon as part of the root. We also wanted something with the first two letters of the alphabet. So when you have alphabetical ranking, you're at the top of lists, right? So anytime you go into a trade show, et cetera, we'd be at the top. So we said Alexicum, Alexicon. Both versions sounded a little cumbersome and we shortened it to Alexa because it was easy to remember, easier to say. We tested in different languages, didn't mean anything offensive. Pretty much every culture we tested it with and every language, people speaking different languages had no trouble saying it.
So we're like, okay, it's a good name. It's catchy. We should call it Alexa Translations. Applied for our trademark in 2012, 2 years before Amazon did in Canada. And we have the trademark for it.
How have your company grown from the early days to now? And you can use any metric that you feel comfortable with, but I really want to understand the difference between these two snapshots and how the company evolved.
So obviously since 2010, we're over 100 times bigger. And so I mean, just the size-wise, you know, just in our core management team right now, we're at over 40 people. So it's definitely a different organization than when I was first doing it myself. And part of the growth happened 2 years ago with an acquisition in the AI machine translation technology space, 2 and a half now. And what triggered the acquisition was that human translation was great and we are one of the best in the country and one of the largest in the country, but we knew that AI was playing a bigger and bigger role. And I think to anyone listening, I mean, they would have noticed with Google 3, 4 years ago, Google all of a sudden started getting much better.
I mean, you go onto Facebook these days and you see translation in different languages. You're like, yeah, that's actually quite impressive for a machine. And as we noticed that, we're like, okay, we either are going to play part in this or one day we may not have a business. And we found a company that was listed on the New York Stock Exchange over the counter. We made the transaction and got them to go private for the transaction. And we're now one, and then they've been rebranded Alexa Translations AI. And we now have traditional translation services as well as you can actually license our AI and machine translation technology.
Does this signify a shift from being a services company to being a technology company or a product company?
100%. I think there's a large portion of our organization now that is dedicated to just that. I mean, we have 10 programmers in-house, which is sizable for, for, for especially when it's something as narrow as this dedicated to translation technology. And the idea is that we have the only tool dedicated to the French Canadian market. We launched Spanish through it as well, Latin American Spanish, but really the core of it is French Canadian. So English to French, French to English. So any Canadian entity that needs high-end French content, our tool is the best way to go. For it. And we have— we've run objective tests, et cetera. So this isn't just me saying we have the best tool.
And we— so the organizations that have translation departments can license it. Lawyers that are doing due diligence on a file and have tens or hundreds of French documents can upload a document, translate it, maintain the formatting, and get it within a minute or so. And they can do that with 100 documents at a time, right? No one could have imagined that even 2-3 years ago.
But is the human translation and the services related to human translation still the biggest part of your business, or has it now been superseded by the technology side and the product side?
It's still the biggest part of our business. It hasn't been superseded yet, but the projection is that it will be superseded very soon.
Wow, so this is very interesting. Before I ask questions about this, compare the client profiles in 2009 and now. How are they similar? How are they different?
Well, in 2009 we used to still work even with a lot of individuals. So, you know, Pulat said I have some documents that need to be translated into English from from, I won't actually give away your country 'cause you challenge our viewers to figure it out. So let's say from source language to English as a target language, and you'll be like, no problem, we'll get them prepared for you. You need to file them with, say, Canada Revenue Services or Canada Revenue Agency. Here's the translation with, along with the relevant declarations. Right now, it's rare that we work with individuals at all. Like it's all B2B and with some of the largest institutions in Canada, some of the largest banks, largest law firms, accounting firms, and energy companies, consumer goods companies, really there's some of the largest brands Not just in Canada, in the world actually.
Right, and I assume the difference between 2009 clients and 2020 clients is volume and the expectation of quality, quality standards. So the stakes are much higher and you get much bigger projects and standards are much higher than 11 years ago, correct?
All of that is correct, and our errors and omissions insurance reflects it accordingly. The magnitude we're required to have now because of sort of clients we have versus what we had back then is very different.
So I'm really curious about the ability of machine translation to meet that standard. You just described an extremely high standard and you just described your company's successful evolution from a run-of-the-mill translation shop to a corporate provider of high-end translation services to high-stakes clients. Where does machine translation fit in there? And is it really good enough to meet that standard?
So machine translation cannot— you cannot put a prospectus or a statement of claim through machine translation, take it out of it and file it with the court or with the Securities Commission. So machine will not provide you something that is client-ready, so to speak. Machine will provide you something, and what machine aims to be is to get to that over 90% accuracy rate so you can get as close to human translation as possible. But the whole point of the industry is Let's make the translator's job as fast as possible, but not eliminate the translator. Language evolves too much, and a lot of it, a lot of the understanding and so on, lies in the hands of the reader.
So machine can do a really good job, but it will soon require human intervention, at least in the foreseeable future. I can't see a path where you eliminate a translator altogether.
Right. So when you describe the technology side of your business, you mentioned that you were licensing your technology or that you developed products based on your technology. Can you go into more detail about what exactly the technology side of your business sells and to whom?
So the largest buyers of our technology are some of Canada's largest law firms, accounting firms, financial institutions, mining companies, insurance companies. And what they're doing is the internal translation departments are, or bilingual users are saying, I'm going to use this, get the first draft, and then it'll make us more effective internally. Faster so that translators don't actually have to do the boring job of doing everything from scratch and focusing more on the wordsmithing, which is what they enjoy anyway, right? It's almost like somebody could go to a litigator and say, this AI will generate a statement of claim for you that does some of the mundane stuff that you need to do anyway. And you just focus on the wording of some of the issues and some of the stuff that is really what matters to the statement of claim.
And that's what the idea was behind this. It's like, let's generate the technology. You can API to all kinds of third parties, resolvers, et cetera, of the world. It's like any, you know, if you're doing due diligence, you could API to translate what they call CAT tools, computer-assisted translation tools that translators use. But our AI platform alone, you could upload documents through it directly. So imagine a platform where you can upload a text, or sorry, you can copy and paste text the way you would have seen with a lot of the current tools, but also upload documents like Word, PDF, Excel, PowerPoint, InDesign files, HTML files. This will translate it and maintain the formatting and get you something that is impressive in the quality of the format as well as the translation quality.
Your company purchased this technology essentially and then evolved it in-house. Tell us more about that acquisition. So what are the origins of this technology?
The predecessor organization had very good technology, but it was focused on the e-commerce space, which meant that any websites using it, this was translating 30+ languages at a given time. And what we found was the, you know, when you're competing doing that many languages, you're actually competing with the Googles of the world and some of the companies that have resources that are immense, right? And we said, okay, we don't have the resource of Google, but we have very specific knowledge in the legal financial marketplace where we could really have a tool that is amazing in this very narrow profile. Except that narrow profile in a country like Canada that is bilingual means there's a lot of clients that benefit from it.
So we specialize into Canadian French and then further specialize in legal and financial. So that's where it's the the best. But even beyond legal financial, do really well compared to other tools out there.
Gary, I had a few conversations with Francophone lawyers over the years, and I interviewed one recently. I spoke with another one on the same show. And one of the things that came up was that there is a gap between French, between court decisions written in French and court decisions written in English. There's this whole body of law that was originally written in French in Canada, and it's our law. Just because it's written in French doesn't mean that it doesn't apply to us, right? And I'm not even talking about Quebec decisions necessarily. I'm talking about some of the Ontario Superior Court decisions, right, or even Ontario Court of Appeal decisions can be in French. There are whole regions in Ontario that are Francophone, that cater to Francophone populations.
So we talked about this gap, and some of our lawyers are fully bilingual, so they bridge that gap for themselves, and actually differences in translation came up as legal arguments in some cases. Tell us about using your technology to solve this problem. And it's not only a one-way problem. It's not only translating from French into the language that we, the English-speaking lawyers, understand. But it's also translating this giant body of English language decisions into French to actually accommodate the linguistic minority. What do you think about this?
So, you know, it's a few different items in what you just said, so I'll dissect it one at a time so that we can approach. So certainly to your point, you know, within Ontario and New Brunswick, et cetera, like Alberta, BC, we have Francophone communities throughout Canada, right? Sometimes we forget and we think of it as almost a Quebec regional language, and it's not. It's, you know, and even to your point, within Ontario, I'm sure we have judges that are far more comfortable writing in their native language than maybe French. And I'm sure their English is great, but if you had a choice, they'd write in their own language. And if they so choose to write it in French, they are able to, and as they should.
And then you would have this, to your point, great decisions that are in French, but when we get the translation, if the translator didn't fully understand the spirit behind it. And to really understand the spirit of a translation, you really need an understanding of the subject matter, of the ongoing proceedings, and sometimes even the intentions behind it. So it's almost like to actually translate the decision, you need to review all the materials that were submitted to the judge before the decision was made to actually have that level of context. Unfortunately, in translation, if we did that, translation becomes so expensive. And so cost prohibitive that all we do is, here's the court decision, translate it.
Really diligent translators will go do that additional research and look for these things, but not everyone does. So as a result, you'll have sometimes a gap between the actual decision and the translation that follows it. The reality is that if somebody is unilingual, so they either speak English or French, How do you look at other decisions and say, is this relevant to me? So this is where tools like AI of this nature becomes really handy because now you can do research a lot faster and be able to consider Quebec court decisions, even if you didn't understand French, because this tool can get you something in seconds, a court decision in seconds, whether you upload as a PDF or whatever it may be.
And it will give you a very good translation to rely on for the purposes of, is this relevant to my case or not? Not ready to submit as an exhibit in court, but it's certainly good enough for the purposes of knowing, is this relevant to my case or not? And it can be a way for, as people are drafting in their language, to double-check, like, okay, let me think how it comes across in English, for example. So that's where this will come in really handy for practitioners, especially as they're referring to decisions and where there's only a French version of it. This will be— it's an extraordinary tool.
I really like this idea that perfect is the enemy of good. And what you're essentially telling me is the first pass is good enough. And the first pass, of course, is done by your technology, by AI. It's good enough to know what the case is about, to know if the case is relevant. It's not enough for some fine analysis or interpretation of the case. You know, sometimes even my brain is not enough on the first pass when I'm reading in the language that I perfectly well understand. So I totally get your point that translation requires requires external sources and external information, information external to the text to achieve higher quality translation. But the first pass is good enough for so many things and your software is automated because it's a technology.
You mentioned APIs. So, you know what immediately occurs to me? Can your company provide French language headnotes to CanLII decisions? Assuming that CanLII opened its database to your company in the backend via an API, what will it take to accomplish that? Because this is going to be huge if it happens. It's going to be huge for thousands of French speakers in Ontario, but also so many French speakers here in Canada.
We could realize that within a week. That's, I mean, that's the beauty of it because we already have a public API. It's just more investing in making sure we connect properly to, to the CanLII infrastructure. And again, with the development team being in there, and not only, you know, and if anyone from CanLII watches this, this is an open offer. Please take me up on it. We'd love to, to help with that. We have to put in a disclaimer that this is being done in an automated fashion, but for the sake of accessibility, etc., it's better to have this than not at all. And I think we'd be more than happy to collaborate.
This is a great offer. I recently interviewed Colin Lachance, who was CanLII's CEO. I will bring this interview to his attention. I'm sure he has still connections with CanLII. And when you mentioned the disclaimer and this notice that the translation or generation of French headnotes, for example, is automated. You know what I thought? I think, if my memory serves me right, and I may be wrong, I think CanLII itself uses some form of automation or machine learning to generate English headnotes. I don't think they actually have the editor muscle like Thomson Reuters to write headnotes for themselves. I may be wrong, don't quote me on that, but I think CanLII— the point I'm making, I think CanLII is technology native.
They get it. That's why everybody loves CanLII because they are just good with technology. So we definitely need to bring this to CanLII's attention and I think this could be an example for the entire world. This is how you make bridges between communities in this country, through translation. And now because of technology, we can do it quickly. You said one week, that's crazy, it blows my mind. That's unbelievable.
And Pulat, you know, I'm going to tie to something that I'm very passionate about because it ties in nicely to this. I mentioned about me learning French before coming to Canada, and then being somewhat disappointed that I didn't encounter it as much as I did, especially in a place like Windsor, you know, where you're by the border with the US and there are Francophone communities there as well, but you're just not, it's not a daily occurrence, right? And then spending that summer in Quebec and truly falling in love with the culture, the language, the history, and you're just like, this is amazing. We have this in our country. And yet, Somehow we're not as accessible as we should to each other.
So part of the investment in the technology, which is now, you know, well, almost at the $20 million Canadian investment overall, pre-us and after us, very heavy investment to what it was to bring it to where it is today. And it's, you know, the beauty of this is that it will make Francophone Canada more accessible to Anglophone Canada and vice versa. I mean, this is what it does for, and for certain things, the good enough of understanding what it is, is that's enough. And for certain things, let's say it's a company that has a fund company that is introducing a fund, but they're like, oh, we're not sure how much money we'll raise in Quebec.
Translation costs are prohibitive. Well, now we can actually make those costs far more reasonable because if the AI does a lot of the heavy lifting, you, the human translation to follow can be done at a lower cost than it would have previously, and thus making things more accessible for the Quebec market or New Brunswick or wherever the Francophone communities that are reading it are.
Unbelievable. This can be a great project. We have to bring this to CanLII's attention. We have to bring this to every publisher's attention, I think. There are other publishers in Canada, in the world, and I understand your technology is flexible. It is probably fine-tuned for Canadian languages, but I'm sure that this idea can be translated into other contexts, right? And I think that's the missing piece, really. At least I think in the legal community, that's the missing piece. And you're a perfect— your company is a perfect candidate to bridge that gap. Thank you so much, Gary. This has been an amazing conversation. I must say, I did not expect that we will have this idea today, and I'm really happy that we did.
That's what happens when people who know what they're talking— I'm not talking about myself, I'm talking about you— people who know what they're talking about come to this show and share their wisdom and their knowledge and their experience. I'm really thankful to you, and I want people who work with our publishers here in Canada to contact Gary. He's a really open person. He'll be happy to hear from you. And hopefully something really good will come out of this. Thank you so much, Gary, for this conversation today.
Thank you for having me, Pulat, and don't sell yourself short even on the CanLII piece. It was your idea, I just filled in the gaps.
Thank you, Gary, you're too kind. I appreciate it.
Thank you again for having me, this has been a lot of fun.