Speaker 1: 0:00

Why don't you tell us your story? What's your startup's name?

Speaker 2: 0:03

My startup's name is Truwind. Yeah, we're using AI and LLMs in bookkeeping and finance. Why should we use Truwind, At the end of the day, for bookkeeping? You want it to be timely, you want it to be accurate and you want transparency. We do all three of those. Two of them fit for AI, your bookkeeper. You're lucky to get about a 15th of each month, more likely the 20th or the 25th, because your bookkeeper is managing five, six, eight different clients. They can only work on each customer one at a time. Technology can work on all eight customers at one time.

Speaker 1: 0:35

Yeah.

Speaker 2: 0:36

So that's how we make it more timely. Number two in accuracy your bookkeeper is often going to ask you the same question the subsequent month, the question that was asked last month. They're going to ask you again because they don't remember. They're managing so many clients Again. This is where technology shines. And then the third one, on transparency, is more of a design philosophy. That's how we just think about how to make sure there's transparency. It's not just a black box, so our customers always know what to expect. That's why our customers love it. That's why we have 50 customers today. How hard was it getting those customers? Hard, and with each one you get, the next one becomes, I guess, incrementally easier. But then the hard part is getting them at scale. So our first customers really came through YC. We launched on Bookface, we launched the community and it resonated with people. They're not happy with their bookkeeper.

Speaker 2: 1:28

And so we launched in January of 2023, as YC went to 23 started, we got a lot of great inbound. Did people have bookkeepers? They already had bookkeepers.

Speaker 1: 1:36

Most of our customers.

Speaker 2: 1:38

I'd say 70, 80% of our customers switched from another bookkeeper over to us.

Speaker 1: 1:42

So you were in YC and people already had bookkeepers coming into YC.

Speaker 2: 1:45

So very few of our customers were in our batch. People whenever people exactly their alumni went through in previous batches. It actually makes more sense in our batch. Do you really need a bookkeeper? You should. You should have one. You could categorize yourself or, like many companies, they don't do it at all. But the ones who are further along post seed A, b stage you got to do your bookkeeping, you got to get an order and when they're not happy with theirs, they come to us.

Speaker 1: 2:13

What do you normally tell customers in their first coming view?

Speaker 2: 2:16

Break it down to the three things we covered earlier. The value of property in your bookkeeper is we deliver it faster. Right now, your bookkeeper, you get out, you hear from them on the 20th or 25th. You'll get ours no later than the 15th. For many of our customers you get out on the 10th, even before that, and with AI we're driving the timeline down. That's the objective. We're not going to ask you the same question twice. We're not going to ask you a question ever. There's certain things only you know. That's where the beauty of LLMs really shine. We need to ask you a question. You're able to respond in plain, simple English and are using LLMs. We're able to understand what you said, but we won't ask you the same question again, and that's how we're able to beat out the competitors. So what's remote? The classic question that NFX likes to ask in this space. I don't believe in. We're going to be in the office.

Speaker 2: 3:04

It's very fitting that we're in NFX's office.

Speaker 1: 3:06

We're in NFX's office, lights shining at us. I got to ask this question, right.

Speaker 2: 3:10

I really subscribe to NFX's breakdown of how to think about modes. At the end of the day, technology is never a moat. It always boils down to four business modes Brand scale what was it? Embeddings. And network effects Embeddings. Embeddings. I think that's the phrasing they use. Brand and scale are not available to us. We're startups. Embedding is like are you embedded into a core operation?

Speaker 1: 3:33

Oh, not like Clip. A core workflow. Sorry, not like Clip.

Speaker 2: 3:37

No.

Speaker 1: 3:37

Okay, we're talking about AI, right?

Speaker 2: 3:41

At the end of the day, accounting and bookkeeping is mission critical. It needs to be done right. So for us it's embedding ourselves into the core operations and workflows of the company of accountants, of financial analysts, and I do believe in FinOps. There are niche network effect opportunities to take advantage of. Example would be payments using billcom. If I receive an invoice, I upload it to billcom and then I'm going to pay you through billcom, which means you have to get on billcom.

Speaker 2: 4:12

And there are little niche areas in the financial operations space to capitalize on network effects, and we can do that at a greater level with LLMs. So that's how I think about most. It comes down to business boats.

Speaker 1: 4:25

What first attracted you to AI?

Speaker 2: 4:28

It wasn't so much attracting me to AI, but more so first attracting me into the world of finance. So before I, this is my second startup. My first startup was a company called Blue Light. I went through YC and went to 21. Before that, I spent a few years as a junior VC. That was my introduction to world of finance. I just have a fascination for how financial markets work, how our financial infrastructure has been built up. So my first company, blue Light, was building data ETLs for specifically focused on finance. What's an ETL? Extract, transform and load.

Speaker 2: 4:59

We're solving data problems in the financing growth categories. Then, when I started truing to focus on bookkeeping as our first wedge into the market, I started keeping my ear to the ground on all things happening in AI. Heard things about GBT2. Heard things about GBT3. Started asking a lot of friends what their experience has been with it, what it can do. And that's when it kind of clicked for me.

Speaker 2: 5:24

In accounting we've been trying to automate accounting ever since JavaScript was invented into the 1990s. At a high level it kind of makes sense. Accounting is just a set of rules. Right, it's only numbers. We should be able to solve accounting with rules based software. But that's the mistake. Accounting is not just a set of rules. If you work at a startup and you buy a cup, that's an office expense. But if I work at your local blue bottle and I buy the same cup, that's a cost of goods sold. It's the same transaction with two different outcomes. That's a problem that software and basic math can't solve. You're missing an ingredient, you're missing an input, and the input being why did you buy this cup? What does your business do? That information is stored in natural language and that's what we finally have the technology to capture at scale today. So we believe with this type we can finally put automation in accounting at the level that we've been aspiring for for decades.

Speaker 1: 6:18

What's been your biggest challenge? Trying to integrate AI into your products.

Speaker 2: 6:23

There's this big question around human AI interaction. That I think is a big question mark, and anyone who says they've solved it I don't know. I think that's a little. Maybe it's on too, is I think it's still early days Like, do I believe in AI should be able to handle all things in accounting? The answer is yes, but not yet.

Speaker 2: 6:44

If I were to ask an AI an accounting question and AI responds, there's going to be some reservation on my part, thinking what if the AI got it wrong? Yeah, I would feel more comfortable with a human. Now it almost seems silly, because humans get things wrong all the time, but we're in this paradigm right now where we're more forgiving of humans to make a mistake than of an AI. That's probably the biggest challenge that Cruz has has far fewer accidents than human drivers, but the moment something goes wrong, it's all over the front page. So there needs to be this gradual rollout of AI in the human AI paradigm that we're accustomed to, and so thinking through that user experience, thinking through those workflows that's probably the biggest challenge here.

Speaker 1: 7:29

What do you plan on training your own LLMs to help you?

Speaker 2: 7:33

Eventually, as an early stage startup, the goals can move fast and so we build on top of the existing LLMs out there, on top of GPT-4 for zero shot, for GPT-3.5 for our own fine tuning and multi-shot. But eventually I think it'll make sense to train our LLM model focused on finance and accounting. I know Bloomberg launched something a few months ago. Intuit is doing their own things, but we don't know what those things are yet. It's still kind of kept in house and when the time is right it's something we're going to be looking into. Just not yet I think it'll slow us down in terms of capturing the market.

Speaker 1: 8:07

Yeah, we thought very similar things.

Speaker 2: 8:10

What's your view on it?

Speaker 1: 8:11

I mean we don't train our own. I mean that's not completely true. We do train our own LLMs, but we mostly train our own LLMs to test our LLM training software. Not because we need LLMs to be trained, but it's fun. That's my view.

Speaker 2: 8:27

You do as you should do, because it's fun.

Speaker 2: 8:28

Why else should we do this right? We've got one shot at life. I should have a good time doing the things you want to do. What made you originally want to become a founder? So people I wanted to work with. I met these two really bright, very talented, hardworking, humble guys. We really got along and we had the same view of how to use technology to solve the data problems in finance, worked together for a few months, had a great time and decided to take the plunge together. So part of it was a desire to go solve in this problem space and the other desire was to go work with people I respected and wanted to work with. Same the second time around, with Truand Tennyson, my co-founder in CTO, has been phenomenal, and so being able to work with them every day we're both here in SF in the office every day it's fantastic. It's so much fun. I want to continue building in the financial operations space the CFO office that I'm so passionate about.

Speaker 2: 9:25

I think that's life's too short not to spend time with people you like, right? So everything boils down to the people.

Speaker 1: 9:32

I can't imagine. It's all been rosy though.

Speaker 2: 9:34

No, definitely not. There's been good months where everything seems to be firing on all cylinders and you're trying to close a certain enterprise customer and then it falls through. I think if we were to take a step back and view the events that occur, it's not as bad as what we may feel.

Speaker 1: 9:55

You know what I mean, yeah.

Speaker 2: 9:57

Like sometimes you get that knowing you're like, oh man, is any of this ever going to work? You know what's going on. You know there was in one scenario there was a very hot prospect, making a lot of good progress, and then things kind of went quiet for a couple of weeks. Turns out, you know, they just had a couple of fires and they just couldn't engage again. And then now we're continuing the conversation and getting really close to hopefully inking something into contract together.

Speaker 2: 10:23

So, I think when you say it's not always, always, definitely not. This is a grind. You have a lot of rough days and got to cherish the good days when you have it. And I think what helps is when you take a step back and look at the at least month over month, not day by day, but across months, push things in perspective.

Speaker 1: 10:41

How do you deal with those rough patches?

Speaker 2: 10:45

Having a co-founder has been. Yeah, it does come back to it. I will say, when I started True, and I was I was unsure. Admittedly, you know, going through a co-founder separation, as amicable as it was, it wasn't easy. So second time around, I thought you know what I know having a co-founder is a good idea. I'm just a little you know, still healing, so to speak.

Speaker 2: 11:12

So I still invest some time to find a co-founder, just not all my time. I'm going to go bill, I'm going to go get customers, I'm going to kind of embark on it with this mindset. I'm going to go out and buy on my own, but still keep an eye out for co-founders, Because I knew that was a smart decision. I knew in a 10 year horizon that is a smart decision. How? A co-founder? Yeah. Fortunately, my co-founders spent 100% of his time looking for a co-founder. I met with seven to 10 people. He met with 70. Yeah, and I was very lucky to be one of those 70. So we lean on each other and he's yeah, he's been phenomenal.

Speaker 2: 11:53

Just great all around.

Speaker 1: 11:54

What were you looking for in your co-founder?

Speaker 2: 11:56

Let's kind of reframe it when you're kind of like looking to pick a friend. Could you break that down into words? I think the answer is yes. If we're doing this job, we should have at least a decent read on people, and sometimes you need to trust your gut instinct. But your gut instinct needs to be well informed. One thing that Tennyson and I did and I recommend this to all people co-founder dating. Over the Christmas holidays last year, I invited him and his girlfriend to come to Seattle and spend the holidays with me and my family. So it was me and my girlfriend, my brother and his girlfriend, tennyson and his girlfriend and my parents, all in a four bedroom house in Seattle.

Speaker 2: 12:35

When you spend 24 seven with someone for a week straight, you really get to know them. We had breakfast together, worked together, played games together, walked around together, spent all day together. You start knowing each other's like. You know hygiene habits, bathroom habits, do they irk you in any way? And those things you should pay attention to. And after that week, like we knew, we got along. We had so much fun just hanging out in the same space. And yeah, without getting into the specifics of you know the tactical pieces I looked at, because I think that depends on your personalities. I do recommend this to all co-founders who are co-founder dating Go spend a week, lock yourself up together, and if you have a family and you can't do that, you know understand that's difficult. Three days, four days, you know Thursday through a Sunday. You really learn a lot.

Speaker 1: 13:24

What did you learn?

Speaker 2: 13:26

That Tennyson and I we liked hanging out together. I don't think you can understand that we just enjoyed hanging out together. We can talk endlessly, you know. One question we were asked was what kind of like shared activities do you have? Tennyson enjoys good food. I enjoy working out. Now he enjoys working out too, but at the time is we can just talk endlessly about technology, about product, about politics, about people, about everything. We shared the same intellectual curiosity, so to speak, and that way, you know, we can sit in a room for five hours and we'd enjoy every minute of it.

Speaker 1: 14:02

What would you do during that time? Just chat, okay.

Speaker 2: 14:04

Yeah, we would just chat the whole time.

Speaker 1: 14:06

Yeah.

Speaker 2: 14:07

There are many topics of conversation that frequently come up Politics comes up a lot, especially as you get into the new political cycle.

Speaker 1: 14:13

Yeah, what are your politics?

Speaker 2: 14:15

Man, I can't talk about that, maybe for another time. Are you EAC? Yeah, yeah, we can leave that for another time.

Speaker 1: 14:24

Did you see that Gary Tan just came out as a e-act?

Speaker 2: 14:27

I have been following, At least at a glance. I just kind of glanced through Twitter, trying not to spend too much time there.

Speaker 1: 14:34

Where do you get most?

Speaker 2: 14:35

of your news. My favorite publications I like reading the Wall Street Journal and it's discipline. I like reading the Wall Street Journal. I really like Axios. They actually covered our funny announcement, lucinda. It's just fantastic. So I follow. I read the Wall Street Journal. I read Axios. I do skim Twitter. Just try not to spend too much time on it. Yeah, because otherwise it gets sucked in. Otherwise I gotta say discipline somehow.

Speaker 1: 15:05

Yeah, there's a lot of nonsense on Twitter too, right?

Speaker 2: 15:08

There's also some entertaining stuff on Twitter.

Speaker 1: 15:09

It's entertainment, it's not just news, it's the entertainment. I like the Wall Street Journal, which is very purely the news.

Speaker 2: 15:16

I love reading the Wall Street Journal. I think it came from my VC days. I think it's the most respected publication. I love reading it, yeah.

Speaker 1: 15:24

When I was a kid, actually my father, every morning he would spend like two hours reading the news.

Speaker 2: 15:28

Really.

Speaker 1: 15:28

Every different publication.

Speaker 2: 15:29

What did your dad?

Speaker 1: 15:30

do? He was a lawyer.

Speaker 2: 15:31

Ah, I see, yeah, love it, but also an investor.

Speaker 1: 15:35

He was always reading the Wall Street Journal and the Wall Street transcript. I don't know how we fucking read the Wall Street transcript. I think it's dense. Wow, he just read a cover to cover every day.

Speaker 2: 15:45

As discipline.

Speaker 1: 15:46

Yeah.

Speaker 2: 15:48

I think that's the picture of technology is there's so many things that we don't know about how AI and generative AI will play out. I don't know. I'm going to have to leave that to people smarter than me. I have a view as to the future of accounting and financial analysis. That's the world that I live in, but as for specifically the future of technology, I don't know.

Speaker 1: 16:11

You think that we're going to have accountants in five years and ten years I do.

Speaker 2: 16:17

When you think about why accountants and analysts exist, you got to take a step back and think why does finance exist? What is finance? I think that dates all the way back to the first time anyone brokered a trade. When I specialize, I'm going to specialize in only farming chickens because I can sell my chicken for something else and I can produce more of them. And then at some point we create this idea of being able to invest. This ship is going to embark on a trail to do trade in the Far East. You can invest in this ship, but three out of ten ships never came, or maybe only three out of ten ships will come back. Dangerous time for seafaring, very limited capital going into it. And then we came up with the idea of diversifying risk. You can invest a little bit into ten ships rather than everything into one. More capital started coming in and really that was the impetus for the first publicly traded company. That's the genesis of the idea. Today FastForce is. Today we have this very vibrant public market, equally vibrant private market, really driven by, shall we say, trust. Trust that the SEC is doing their job at providing the right oversight. Trust that the accountants are following the rules and regulation and doing audits and checks in the same in the right way. Trust that financial analysts are projecting in the right way and equity researchers are measuring that appropriately as well. Convey that information to investors. So accountants they serve as such an important backbone to compliance and how this financial infrastructure works.

Speaker 2: 18:02

Could you take humans out of this entirely and trust it entirely up to AI? Maybe, maybe. So the foreseeable future. I believe technology, so maybe this goes into where I see the future technology. I see it as an enhancement to humans. I really do not see it as a replacement to the jobs that we have. It is a question of how can we use this technologies to make ourselves 10x better, to unlock the creative minds that maybe we could unlock, because we're spending too much time doing the nitty gritty data entry things. So, to touch on both, when I see the future technology, it's the pair of this technology, pair this AI co-pilot with every what I call a professional. Make us not just 10x more productive but really 10x more creative, and I think a lot of new things could come out from something like that.

Speaker 1: 18:51

Is there anything you would do different in your next?

Speaker 2: 18:52

company. I learned a lot from my first time to my second time right Cause I have to benefit a hindsight. Yeah, what do I learn from the first time? The importance of thinking about distribution from day one, not just thinking about technology. Having the courage and the nerve to ask someone to pay for something when nothing exists yet Really, if it is just a proxy to see that they'd be willing to pay for it. Knowing how important speed is at the end of the day, nothing matters except speed and storage. You have to be willing to move fast and be willing to take chances in it. So I learned all of that from my first starter. Going to this one made sure my co-founder and I got along. We had a clear vision and strategy, move ahead, got customers from day one to be design partners and work with us in developing this, and always being creative on go-to-market and growth channels. As for learning so far, well, jury's still out. We're trying a lot of things. Dvd if it's gonna work or not.

Speaker 1: 19:54

Yeah.

Speaker 2: 19:54

But a couple of things have worked. We're up to 50 customers.

Speaker 1: 19:57

Yeah, I wanna be at 100. Would you rather have a small number of very high paying customers or a lot of medium paying customers?

Speaker 2: 20:08

What is that drawing there's like the nat, the mouse, the deer, the elephant and then the whale.

Speaker 1: 20:15

Is there a whale? Is there?

Speaker 2: 20:16

one above elephant I can't remember 100 now. Our customers are in the deer category right now. I think the answer to your question depends on the space that you're in and which customers you're going after. For us, accounting and finance, it's a very delicate thing.

Speaker 1: 20:33

Yeah.

Speaker 2: 20:34

Can you imagine me knocking on the door of Nike and asking them to handle the keys that came them? I can kinda see it.

Speaker 1: 20:38

I think I can see you.

Speaker 2: 20:40

I guess I'm wearing all the right gear, but it's gonna be a hard sell because they can't afford to get it wrong. Yeah, not to mention the CFO. There's this perception that, oh, this office at the CFO is cost cutting and so they rarely adopt new technologies. All that together we thought, okay, we need to go in lower market first, let's go to SMBs, let's go to startups. We close seed stage customers. That gave us credibility and enough. What do we say? Social proof? Yeah, we get Series A customers, then we got Series B customers, then we got Series C customers. So our mindset is this tool deserves to live in the hands of accountants and financial analysts. They exist in larger categories, in the later stage companies, so let's keep pushing up market and getting our tool to them. So the answer to your question really depends on what category you're in. For us and makes sense to go for the deer and the elephants.

Speaker 1: 21:36

When did you first decide that you liked finance?

Speaker 2: 21:42

Probably when I went to business school. So I was an aerospace engineer at the summer career. I designed and built airplanes at Boeing for five years. Well, I spent three years in product development working on the 777X and then got to sell the airplane to Cathay Pacific and Singapore Airlines particularly special because I grew up in Hong Kong. So if you want to sell the Cathay, it was pretty darn cool. Helped sell 20 airplanes to Cathay and 40 airplanes to Singapore. Then went into went to get my MBA at Columbia. I was one of I think I was the only aerospace engineer in my class class. Above me there were two. I came in. I went into business school with such a fresh, I guess like what was that Eyes? A bushy, bushy tail.

Speaker 1: 22:25

Eyes wide open.

Speaker 2: 22:26

Yeah, something like that. Bushy eye, bushy eye, yeah, yeah, yeah, no, no, no. Bushy tail, wide eye, there we go. Yes, and just was fascinated by how the world of finance worked, how financial markets worked, how this public stock market attracts so much activity and how people operate on such a high level of trust.

Speaker 2: 22:50

Then I found my way into venture capital and saw the dichotomy between the two. On one hand, you have private markets. A transaction happens once in a while, usually happens behind closed doors. There's generally no liquidity. I think there are more and more secondary companies trying to figure this out and I think that's fantastic. I think it needs to but there's no liquidity and no information disclosure to. When you go public, it's a step function to 100% liquidity and 100% information disclosure.

Speaker 1: 23:21

Yeah.

Speaker 2: 23:22

Why can't there be something a bit more gradual, so there can be more liquidity along the way, so people can participate in the wealth creation that happens in startups? I know we can't, because of the way it currently exists, the SEC doesn't have enough oversight into it, it doesn't have the resources, so we form an idea called credit investors. That can only participate, but I think with technology, with AI, we can do something far better, and so that's you know. It's the fascination with how the financial infrastructure creates so much wealth and from the wealth, creates more innovation, which creates more wealth than the financial infrastructure being the backbone to all of that. Yeah, that was the curiosity that drove me in this direction and has me building on this career path.

Speaker 1: 24:03

That is a great backstory. Thanks, Finn. I feel like you went like an entire transformation in this story.

Speaker 2: 24:10

And I started off building airplanes. My parents still think it's the coolest thing I ever did. My dad worked in the satellite industry and rock industry. So when I worked at Boeing, oh, he also brought my son works at Boeing. When I went to the venture capital, I think my dad started saying my son does something in finance.

Speaker 1: 24:28

I mean I'm pretty excited to be in the room with somebody who worked at Boeing. Thanks, man. Did you ever fly any of the planes? I haven't got my PPL.

Speaker 2: 24:36

I've one of my closest friends. He's a captain in the Air Force OK, now out E. I think he's a flight instructor as well, so I've flown with him a few times. It's fun.

Speaker 1: 24:50

In these planes. Yeah the same role In the plane that you designed.

Speaker 2: 24:54

So the plane I designed has flown, but it hasn't been certified commercially yet. Okay, yeah, should have been last year, I think, or this year, yeah.

Speaker 1: 25:03

How does it feel to design the plane and then not get the fly in it Soon?

Speaker 2: 25:07

Yeah, Soon, We've. I don't know where they are on test flight, but it will be really cool. One day I'm going to hop on a flight to go back to Hong Kong to visit, visit family and the triple seven X wingtip is going to fold for takeoff.

Speaker 1: 25:26

So there's a folding wingtip.

Speaker 2: 25:27

So this is a really cool thing. They enlarge the triple seven X wing. So much from its predecessor, the triple seven, three inch ER. They enlarge the wing to increase range, increase fuel efficiency. But the wing got so long that it fell into seven four seven category, which means this new airplane would have to park in a seven four seven parking lot or gate at a airport, which becomes a logistical nightmare. That was unacceptable and so they decided what do we just fold the wingtips. So after a land you're going to see this airplane the wingtips fold up, it's going to park and when it takes off.

Speaker 2: 26:04

it's going to fold back down. Yeah, it's going to be really cool. We've had this tech for a long time in the military and we're going to use it commercially now.

Speaker 1: 26:11

Oh yeah, I think I've seen planes on.

Speaker 2: 26:14

that's how they fit, that's how they fit all those airplanes on the carriers. Yeah, hold it up.

Speaker 1: 26:19

It would be a little bit nerve wracking seeing my wing just to fold away while I'm getting on the plane.

Speaker 2: 26:25

I wonder how they're going to introduce that, because I can imagine as a consumer, you look out the window and think where's the wing? The heck. I hope that doesn't happen while I'm in the air. There was a lot of design thinking that went into it to make sure nothing like that would happen.

Speaker 1: 26:39

I mean, you know, this is actually why I don't paraglide anymore, because I did, because the wing can just disappear while you're flying around. Yeah, that's scary. We're even supposed to practice it when you're going for your paraglider license. You have to stall. You have to practice stalls. I'm like what happens if it doesn't untangle Nerve?

Speaker 2: 27:03

wracking.

Speaker 1: 27:03

Yeah, you've got an extra little paraglider that you can throw out.

Speaker 2: 27:07

Do you have you ever skydive? I have not.

Speaker 1: 27:12

I like having a wing. I used to fly 160 ones tiny planes, and there you have a wing.

Speaker 2: 27:23

I have never skydive either. My excuse is, guys, we built this thing so you would never have to jump out of a plane. You're also insulting my profession by jumping out of it willingly.

Speaker 1: 27:35

It is one of the progressions, though, on the way to wingsuiting, and wingsuiting looks awesome, appealing, appealing and terrifying and a little bit tempting.

Speaker 2: 27:43

Yeah, I imagine that would be a little bit cool.

Speaker 1: 27:45

I think it takes a lot of practice.

Speaker 2: 27:49

Maybe in my next life, next startup, yeah, so I think that's it for time.

Speaker 1: 27:57

It was really great having you on the podcast. Awesome, seeing you again, man.

Speaker 2: 28:00

It's been a while, yeah, I've been seeing you. Gotta come out to SF more often. You, you, you, you, you, you, you, you, you, you, you, you, you, you, you, you, you, you.