Transcript: How Yoodli is Replacing Boring Sales Training with AI Roleplays | Varun Puri, Co-Founder & CEO of Yoodli

Host: Hello everyone, welcome to startup project. Today our guest is Vrun. Vrun is the co-founder of Uli. Uli is a Seattle-based startup that currently valued at $300 million. recently raised their series B in December and raised their series A, if I remember correctly earlier in last year.

Host: Um, Uli is an interesting company. they had an interesting journey. They started as a B2C company, then pivoted to B2B, uh, I just found their product market fit uh in B2B space. Uh, so I think it'll be a fun conversation about how they evolved.

Host: Um, and what are the cutting edge products that are coming out of AI space? because I think with AI, we are seeing products, new type of abstraction that was coming in and built built on top of LLMs and I think Uli is a good example of like leveraging AI and building new type of products.

Host: Uh so if you are interested in new cutting edge AI products that are being built. I think this is a good conversation to hang on to.

Host: Uh with that, uh Vrun, welcome to the show.

Guest: Uh, thank you. honored to be here. I was telling you just as we started, when I just started Uli, I was watching one of your episodes with David Shim at Red AI.

Guest: So I'm honored to be part of this.

Host: Yeah, two I think uh Seattle-based uh companies that uh sort of uh making ways in the AI space. Both Red AI and Uli.

Guest: The fact that we're even in the same wavelength and story book is read is awesome because anytime I hear about Red AI, they've raised a gazillion dollars, they they're ARR is a Brazilian. So we are by no means close to them, but the fact that you use us in the same sentence is exciting.

Host: Yes. uh let's let's start with the early days of Uli, like what was the initial thesis of starting Uli? Talk a little bit about, you know, what you guys tried before, um, you know, what it became today.

Guest: Yeah. Uh great question.

Guest: The thesis is still the same. It's so Isha, my co-founder, she was at Apple, I was at Google. Uh, both of us just saw so many smart people miss out on opportunities because they don't back themselves when speaking.

Guest: We saw this in big tech where the person who deserves the promotion doesn't get it because the smooth talker gets it. We saw immigrants and introverts and non-native English speakers, uh, miss out on promotions.

We know so many people who struggle with their public speaking. You know, we've all been there the night before a speech and interview as we are pacing back and forth talking to a mirror, a camera or a stopwatch.

Guest: So Uli's vision is very simple. It's how do we help people find their voice and confidence? Uh, be the grammarly for speech, the duolingo for speech, the Strava pilot on Apple Health for speech. I'll give you the fast forward story and then you can you can drill in but so started Uli as an AI powered public speaking coach focuses B2C.

Guest: Um you know my why is I want to build tech to help kids in India speak with confidence. We didn't learn public speaking or communication skills. I did CBSE for those of you who know it's very much memorization based.

Guest: Um from there users started using us for, I have an interview, I have a date, I have a stutter, I have a list. I'm nervous before a certain conversation, a salary discussion, a sales pitch.

So we expanded from AI powered public speaking to AI powered role plays, still very much in the consumer space and then we got pulled into the enterprise where now folks like Google, Data Bricks, Snowflake, Ring Central etc, use us to train mostly their GTM, their go to market teams at scale using Uli.

Guest: On Google, I need to train all 15,000 reps on Google Cloud. Uh how do I make sure that they know the right pitch and they can handle objections, use Uli for this. So we've expanded into B2B. I won't say we've pivoted. We still have a consumer product that's still where my heart is and the B2C and B2B reinforce each other.

Host: Before going further into the B2B evolution of the product, what was your learning about public speaking as a skill? Like what are the one or two learnings that are uh you know, sort of like stuck with you while building that product?

Guest: It's uh it's funny you ask. You know, when I was at Google, I became a speech coach for a lot of the execs. So before they go on stage, me, the little guy, I used to record them and make them watch their video back. And me because you know, my judgment didn't matter. I wasn't an executive, I could say whatever.

Guest: Think two interesting things. One is there are just a few rules about public speaking. Speak in threes, start with a hook, have a call to action, use alliteration.

So you know, in terms of speaking in threes you might want to say friends, Romans, countrymen. Let's do it, finger licking good, impossible is nothing, judiciary, legislature, assembly. You get the the gist of it, right? Like things just sound good.

Host: Just broke it with asking you two instead of three.

Guest: Yeah, yeah, yeah, I know, but uh I I can give you three. So that's a trick or when you speak, speak in alliterations, right? At Uli, we hire for culture, confidence and capability, whatever, like when I'm saying the seas, I'm sounding smarter.

Guest: Um so there there are like 10, 12 of these tricks. And then the second and and my real insight is the single best way to uh improve your communication skills, not just public speaking is to record yourself and to watch yourself and to cringe and to do that again.

Guest: We we know that. It's just people hate practice. People hate the sound of their own voice. You know, when I look at myself on camera, I'm like, oh my god, my haircut is like this, all this pimple, I sound like that.

Oh my god, it's too weird. But if I just practice and I do it three times, I'm bound to be better and I'm bound to think more about what I'm going to say, so I'm likely going to impact the audience more.

So our aha, I don't know if it was an aha, we'll see, you know, where Uli goes is, can we just build a system that motivates you to do that.

Guest: So we, you know, we give you feedback on your public speaking elements, your rubric, your content, but all of that is nice to have. You just need to do it three times. Uli is a platform that helps you do that and you're bound to get better.

Host: So is it like uh Uli is could you say is it like duolingo for uh public speaking? Is that the way to describe the B2C part of the business or the product?

Guest: Uh I think that's a good way of describing the first version. The first version was maybe a duolingo for public speaking. Now it's you know, a lot more expansive than public speaking for any kind of uh critical uh communication, conversation.

Host: So what have we guys learned about you know, public speaking is also related to storytelling and I think I I'm spending some time here is because when everything for everything we do, AI comes and does a lot of our tasks.

I think there are only a few meta skills that remain um sort of what differentiates between a high performer versus, you know, an average performer.

I think there are some meta skills that are going to be part of it and one thing that I still think will be valid is storytelling.

Host: Like what aspects of storytelling you know, you have to sort of imbibe into your public speaking. Is there anything that you guys discovered through Uli?

Guest: It's a great question and it's also something that I grapple with all the time. Right? In a world in which AI is going to take most skills. AI will make access to information commoditized. Everyone will know the right answer all the time. Right? At their fingertips.

Guest: At your point, what I what I think separates a high performer from someone else in their personal or professional life is how we show up and how we communicate. What makes us fundamentally human in a world in which everyone has the information and and knows uh what to say is your authenticity, your vulnerability, your relatability. I don't think AI can or will be able to replace that.

Guest: So one of the things I feel proud of is there are so many companies building AI to replace human jobs or to make humans irrelevant. Our thesis at Uli is very much how do we build AI to help humans be the best versions of themselves?

Or in some ways to give humans superpowers against AI. And the irony is the thing we are working on is the piece of being most human, which is how we communicate.

Host: I think that um one of the things that I think eventually will happen is pretty much you see this happening whether if you're like creating marketing content, whether you're let's say ideating features to build a product.

What you're really seeing the differentiator is being sort of like does this person have the judgment to differentiate between what is good versus what is bad because you give an idea now you know, AI can probably give you uh how to execute it.

It will probably execute 80%, 90% of the way.

Host: Um, and we still need to develop that judgment uh whether this is a good or bad.

And I think that's sort of like the problem we'll all face in different parts of the work that we're doing is like how do you develop that judgment in an AI first world? Uh, I think that's some that's a pattern that I'm seeing.

Um, but I think let's pivot to um, you know, AI role plays and your the second product that you came up with or um the I was following your LinkedIn and you obviously write a lot on LinkedIn.

Host: Um and why I think that's how first I got to know about the case study that you published with Google. Um about how they're using Uli for their GTM teams. So how did that happen?

Like did you guys think, okay, we have this consumer product, I think, you know, it might be a good fit for certain GTM roles and you approach Google or you know, you just started experimenting and developing a B2B platform and then approach Google like how did that process actually happen?

Guest: It's a great question also. I laugh when you make the LinkedIn reference because my friends make so much fun of me. I'm that obnoxious loud showy guy on LinkedIn and I'm so embarrassed of it. Like it's not who I am as a person.

I remember in my when we just started Uli, I used to see all these people on LinkedIn just you know, pumping their their chests and talking about how amazing they were and I was like, oh my god, you are such a try hard.

And now I've become that person. So I apologize to the world.

Host: Okay, then then we should actually spend some time there. like what what you as a founder get out of LinkedIn. I think you do it. there is some kind of thinking behind it. So let us explain that thinking before going to the Google.

Guest: Okay, yeah, it's a good question. It's so one, I find it cringy for sure and I and I do it. The reason is I'm realizing social selling is such a big part of this startup game.

Guest: Even if people don't engage with your posts, they've likely seen it. You're likely top of mind for them. It's our number one source of leads online, it's number one source of candidates.

Um and in some ways like my brand is now tied to Uli ideas in the market. So I post three, four times a week. lots of people make lots of fun of me.

But I've realized in this world in which things are moving so, so fast, having a mouthpiece where you can be authentic and quick, uh is very relatable.

Host: And it's also like this this shift also happened where the company or the brand pages are actually if you post the same thing with your brand page, I don't think it has the same effect. Um, right? Like I don't think you could build a lead generation on LinkedIn just using your brand uh page or your company page as sort of like your mouse piece. I don't think that is the case anymore.

Guest: Yeah, it's we've experimented with it and and we've seen the exact same thing. If I post something and Uli post something, whatever I've posted will likely get more traction. It's because people are drawn to authenticity.

Even though I'm loud on LinkedIn, one of the things I try to do is not just be like, oh my god, here's our case study we rock. Uh yay, Uli did this. I'm also trying to share our learnings and our story along the way.

For instance, holy crap, I'm a first time founder. Now I'm managing skip level reports. I just made this mistake. Is this common?

Guest: Or do we hire a seasoned executive at this stage or someone who's fit for our rule? like the kinds of questions I'm asking about management, leadership, how to build a CS team, how to build CS structures?

Do you build tiers in your sales team? I'm sure most founders are thinking about. Do I leave Google to start my company? But what about immigration? How do you think about comp? Like how do you convince the first person to join your team?

I've realized just being vocal about it makes it more relatable to other people.

So obviously selfishly, I hope we get leads and what not, but I've had a lot of people reach out saying, I've had the exact same question and I don't know why people don't talk about this more often.

Guest: As an example last week I posted about, hey, I just got an executive coach and I don't know why executive coaching is kind of taboo like therapy. Here's what I'm working on with my executive coach. What are other people doing it?

So many people reached out saying, oh my god, I'm working on the same thing or I've been too insecure to work with the coach and I'm like great, maybe uh I helped some people.

Host: Yeah, that's I think now your friends you can share this clip with your friends and said this is why I do it.

Guest: No no they know me, they'll still make fun of me and that's what friends are for.

Host: That's what LinkedIn is for. You still have to post that there's an alternative. Um but yeah, let's go let's get back to the Google question. So how the whole GTM case study for Google happened and how did that sort of product evolution happen?

Guest: Yeah, it it's actually been very interesting because at the core we started B2C and individuals just start using us, right? A lot of Googlers, maybe because they saw my posts, started signing up and using Uli because they have a product update meeting tomorrow or they're presenting to their manager or their team and they just start using Uli.

Guest: And through that we had a pretty big ground swell at Google and some someone on the sales enablement team just reached out saying, we need to run this training.

I don't know if Uli can work, but a lot of us are already using the tool to practice our own speeches. Can you customize it around Google's methodology and roll it out? And at the time we couldn't, but obviously my answer was 100%.

This is our Bullseye use case as any founder should say.

Guest: So that started the Google roll out where we added the first pass of an admin interface where the admin can customize what kind of feedback you're getting and that created the entire ecosystem of Uli for admins and for end learners.

The B2C experience being very similar to the B2B experience from a learner standpoint, but integrations and enterprise grade privacy and security and customizations being in the B2B version.

Host: So how does how does the product actually work? Like if if I'm a if I'm running a 1000% GTM team and I'm looking to adopt Uli, what value uh does the company get out of it?

Guest: Yeah, our top line is faster ramp, uh more manager time saved in terms of amount of coaching done and then better win rates.

So what this means is in the old world, if I am, let's call it Data Bricks, and I need to train 5,000 reps on the new Genie product that Data Bricks has just launched.

I would either do this by organizing several seminars and webinars for sellers to attend. I would ask sellers to watch some videos or read some PDFs and at most ask the sellers to do a multiple choice questionnaire.

Guest: Any of those things would take weeks if not months to implement. That's one. Um, I don't know if the sellers actually learned the concept or not because doing a multiple choice is very different from selling the product.

And I don't know if this actually leads to win rates or deals being closed. That's the old LMS learning world where learning was very one-directional and passive.

And our insight has been within seconds, data bricks you can come in and now create a really realistic role play which is I'm a seller, I'm selling to this persona. Let's call it a head of security or infrastructure and IT at a certain company.

Here's what they care about. sellers to prove that you've learned the concept, we want you to be able to do customer discovery, pitch the Genie product, talk about why we're better than competitors and land a follow-up meeting.

Guest: So they will so the data bricks admin will go in and create this pretty incredible role play and it can be very very detailed where maybe you need to share your screen and the AI will ask you questions based on your deck.

And maybe you need to do a live demo and the AI will coach you through that demo. Maybe it's a multipersona role play. So you need to practice pitching the data bricks product to an entire buying committee.

So what someone like a data bricks would do is create a string of role plays.

Guest: Role play number one is where the AI goes back and forth with you and says, hey, do you understand this concept? Let's talk about it. Then role play number two is where they certify you and so forth.

And a seller does this, it's fun, it's this private batting cage where they get better. And I as a leader can then say, oh my god, 95% of my reps are ready for the field.

So you know, I'm taking that like four week, five week, six week ramp and bringing it down to a day.

Host: And over time like do you see sellers actually creating role plays that help other sellers and sort of like create an internal role play marketplace?

Guest: I haven't used the word role play marketplace, but yeah. Right? Like a seller uses this and loves it. And in fact they'll take it home and they'll tell their kid, hey, you know you're applying for a job, use.

They'll tell their spouse, why don't you come and practice. I've never seen people take a sales tool and make it their own.

Uh, we have managers creating role plays, assigning it to their teams and now obviously all of our role plays are agentic where you've just had a call based on that Uli smart enough will say, okay, your next call is with these five people.

I've automatically created a role play. I've nudge this the team on Slack. You're good to go. The manager has an update and so forth. So it's continuous learning, continuous practice.

Host: It's almost like we there's a new abstraction for continuously learn product and understand it and continuously keep yourself updated, but now you're doing within a completely new format than you know, previously it used to be just you create a small learning 101 for every new feature you created and that's why we used to have this learning management systems and used to put a course. everyone takes a course, how to looks at a five minute video and then you know, goes through multiple question and answers and that's how they sort of proof that they've taken the learning.

But I think this is probably more interactive way.

Um do you see like AI role play as a phone factor to sort of you where do you see this going? because in some sense role play we imagine like you know, I imagine role play initially like there's an AI avatar talking to me.

Host: Like that's how I imagine then which I feel like yours is slightly different. Um so are we getting there or like how do you see in general the concept of role play evolving?

Guest: Yeah, to be clear, our role plays are based on avatars, right? My customer avatar will talk to me, go back and forth with me. But that's just one modality.

I think in the old world, learning was one-directional, now learning is conversational and bidirectional. I think AI role plays are just a start. The category we are building is experiential learning. Um this goes way beyond sales and go to market.

Imagine I'm going to school in India, right? and I'm learning about the heart, rather than me reading a textbook, then doing some paper, instead now, you know, Varun AI is like, hey, here's where the heart is in the human body.

Do you understand and here's where the kidney is. Varun pointed out to me. Say that again. Okay, great, you've got it. Now let's get into exam mode.

Guest: And rather than it being a written thing, it's talking to me saying explain the functionality of the heart. No, you got that wrong, say that again. So it's we are making everything conversational, the same way a human would explain it to you, now the AI is doing that with you and then the AI is certifying you.

Host: Got it. Can you talk a little bit about, I mean obviously you guys raised series A then series B last year. So how was the growth has been? Like like what is the customer adoption like been?

Guest: Yeah. Uh so it's been an exciting couple of years. Revenue grew 900% plus uh in the last you know 12, 14 months time period. We have we're closing in on over a million folks using the platform. Uh many large Fortune 100 folks are using us.

I mentioned some folks, you know, Google, Ring Central etc, but also Snowflake and Salesforce uh in the tech world and then large banks, large Pharma. The team was 10, we are now 55 people.

So it's an exciting time. it's also time of a lot of learning and a lot of mistakes, so we're getting better.

Host: Let's talk a little bit about like the AI LLM and like models under underneath Uli, right? Talk a little bit about how that usage changed. Like previously when you were just pure B to C, what were you using behind the scenes and now you know, two, three years later like how does that look like underneath? What is powering uh Uli?

Guest: It's a good question. The one word answer is modularity. When we started, we were obviously on Open AI, the whole universe was. Now we use every LLM essentially. You need to use Gemini for Google use cases.

We need to use Anthropic elsewhere, we need to use Open AI on Azure elsewhere, that's fine. So it depends. We want to meet our customers where they are. And by this I mean across all parts of the stack.

If our customer lives on Slack, Uli will integrate into your Slack. If the customer uses an LMS, Uli can live inside of that. If you just have a website, Uli can live there.

Going downstream, if a customer wants us to use a certain LLM, we can just plug and play. Uh so yeah short answer is all of the above.

Host: So how much of like new product thinking is coming from like because LLMs are improving?

Like does that automatically show up as a feature? because there's always this like I think three days before Gemini introduced making music within Gemini chat, right?

We have an exclusive like a billionaire company called Suno AI which does that previously, but right now it's within the Gemini chat. I can just type what I want as music and I can get it. Right?

So the sort of counter point of view for role plays in general, you could make that. Okay, what if like Gemini makes it too easy? Um you know, to just have a role play.

Uh what if just I, you know, tape all my dry files and just have a perpetual chat where I can just continue to keep, you know, role playing for this particular scenario and just do that repeatedly.

Like that's probably a lot a question that comes up often whenever we're talking about like companies that are are products that are being built on uh LLMs. So what's your thinking when you when you're opposed with such a question?

Guest: It's a great question. So there are two parts to it. One, what drives feature set? Is it LLM improvement and two, what happens when the big dogs build a version of this, right? It's only a matter of time. So part one is very simple.

It's it's all driven by customer need. If an LLM has certain improvements, of course that'll be in the Uli product. Multiple accents, multiple voices. We are not building the iPhone again from scratch.

We are not the ones doing the underlying avatar technology. We will take best in breed that's out there, accent detection, whatever it might be.

Uh LLM functionality we'll include in the platform, but everything on our roadmap is driven by customer need.

Guest: Uh no matter how big if it's like skim, scorm, enterprise integration, all the way to make the role play more back and forth. To your second question and I'll expand this a little bit is everyone is building role plays.

I think we were first to market with it, but it's going to be commoditized, there's no question. Every LMS, every CMS, every CRM, every LLM will have a version of this.

The way we think about it is role plays itself are very much a wedge, but this whole experiential learning play is a much bigger category and not feature, right? It it involves deep personalization and customization.

It involves an entire stack end to end. Creating a prototype of any Gen AI application today is so easy.

You can create a Google prototype now, but all of the underlying infra to build, maintain, support, customize like deploy at scale is where the company building comes into life.

Host: And the big thing, the security and compliance, that's another big thing.

Guest: Of course. I mean but look, security compliance can't be a mode, but it's more like just how deep can the interface be.

So we often say, you know, we're using an LLM, then we'll customize it based on the training methodology you're using. then we'll customize it based on the company you are. then we'll customize it based on the learner you are and then we'll give you a killer UX on top of it.

Now if Open AI were to go after this with this much focus, would I be nervous? 100%. I don't expect them to go so, so, so deep uh into GTM enablement.

Host: In some sense then for the argument like okay, um I think we are seeing right now is um you know, most of this SAS companies which have 85% margins will not have the same margins or like people will wipe code this.

I think you're almost given an answer for that is even if you wipe code, you cannot go that deep. You cannot make all the integrations.

You cannot deeply customize it and then maintain it and then support it and then you know, perpetually improve it with whatever latest comes. Um like do you think about that aspect of this whole phenomena?

Guest: Totally. I think I get questions about what if I wipe coded? I'm like great, do it. that's and and that's my biggest learning in building a company.

Coming up with an idea is the easiest part. building a prototype is the second easiest part. Finding customers and solving someone to use it is the hardest part. And not just building an entire ecosystem around it.

I remember when I was the first when I just starting you like, I can't tell you my idea, it's too good.

Guest: Now, I feel embarrassed. I'm like, if someone wants to steal my idea, let them run with the idea that like that's 0.0001% of the process.

Um so yeah going back to your question, I think can folks wipe code, you know, parts or functionality of any other tool here and there, but uh building something sustainable requires so much more depth and is years worth of work.

I wish I could wipe code Uli in a in a day, it would save me a lot of heartache.

Host: I think the other aspect of this like the economy of scale aspect of any company, right?

One of the reasons like uh you know, a lot of free products that we get from let's say Google or um our cheaper products. like let's say for example take an example of Substack.

Uh if you try to build your own newsletter on a open build your own SMTP server and host it on AWS and just if you're just doing that for your own newsletter, I'm guaranteeing you to get started you will have to spend somewhere between $50 to $150 with zero subscribers. no one actually reading your newsletter and then you have to build trust on your domain uh before actually, you know, making sure that it delivers to the readers if you have any readers.

Host: And then you can do that free for sub on Substack. Um and you know, accumulate newsletters to become big enough where it might be worth doing all that painstaking process for yourself.

Um so one, I think people often miss is like you cannot wipe code the scale. You cannot get the advantages that you get from a scale, right?

If you have for every like SAS company which has thousand paying customers, that like 100k to 200k free customers that are just using the product within certain limit. You see pretty much this with every SAS product out there.

Um so that is sort of like the economy of scale. But I think I also think there is also like a shift happening in terms of like how you do things. Like yours is a classic example.

Like every previous the LMS was the method to share information and learn and that is evolving, right?

So one way to think about it is those companies have to evolve to do it better and while you know, there's always like we had outlook and then slack came and changed how we communicate sort of angle to every product.

So in in some ways it is a natural phenomenon that we see in general.

Guest: You know, um I'm coming up with this on the fly but I'm pretty excited about this idea so I'll tell you.

You know, in typical product thinking they talk about we got early adopters and then innovators, laggard, blah blah blah that's the curve.

The way you describe this curve of like I can build a prototype, I can wipe code it and then I go go go and then I become best in breed, like deeply verticalized.

But then all sudden an incumbent who's building an LMS and I can also add a bunch of role plays. It's basically the same kind of curve where yeah you can spin up a prototype that's easy. Yeah, you can build a check the box features and incumbent.

Our hypothesis and I think this is why journey AI application companies exist is across this curve like best in breed always wins. You should do like a post. that's pretty you know what I mean, right? Like it's

Host: Yeah, it's basically a you know, the earlier graph of uh you know, who was the earlier adopters was.

Guest: Correct. When you have like maximum focus and you've got critical mass of users and you are not far too broad, you are likely to be best in breed. And that's why vertical is the winner.

Host: You but what you're also describing at the other side of that car was basically inertia. The inertia of you know, having a paid subscriber base, the inertia of comfort, the inertia of like hey I made what I have to make already. so I'm not interested in you know, changing things the way they are. like that's always there in any system.

Guest: Truly. there's inertia on both sides of it, right? Like when you're best in breed it's like going over the hump of getting the user base is hard. but then the the question is how big is the dam on the best of breeds? Can you just expand this or do you need to go down the curve yourself and start adding additional features?

Host: Yeah, we we have to probably like spend more time on the where how I need to define it better. Let me let me think about it and send you a little diagram.

Host: Yeah, um so what are the in your perspective, you've grown quite a bit. um last year has been pretty exciting. If you have to say what are the one or two challenges for you right now in the immediate what would that be in your point of view?

Guest: The it's just one with the we're seeing demand for this in the market, there's no question. This category is coming to life, no question. It is tech stack consolidation, right? Every buyer today wants everything living in a single ecosystem. I don't want to pay for 500 tools.

Guest: And Uli is another best in breed solution, but I have a best in breed solution for messaging and a best in breed solution for call recording and I'm salesforce. I can just come in and say I offer all of that.

How does a company of Uli's size and stage continue to innovate and be best in breed. Best in breed is not just slightly better. You've got to be at least 60% better than the alternative so you don't get bundled.

And how do we still make it delightful for our buyers who are buying other pieces of software. So the way we are doing this is integrations, flow of work, make it really simple, but that pressure is becoming more and more real.

Host: Is that like I mean, this has been also a side effect of white coding, right? Because everyone's perspective is that okay, when we can code so many things in the way we like, why should we have so many SAS vendors in a single company, right?

An average like public company has it like I don't know thousand SAS tools on their parole.

Uh, so I think that sort of is also a challenge that new AI products have to cross. not just you but like everyone has to show a much more value than you've shown previously because they expect the user expectation has changed because of white coding, right?

Guest: Without a doubt. not not just that. I mean you know you're in the VC world as well. The VC perception has completely changed, which is the barrier to business on all fronts is near zero right now.

Which is to prove actual value you have to ironically put in a lot more work not less because what was your mode maybe in the early days you had certain connections, you had access to something is now all democratized.

There's no running away from is this something your customers use and is this something they use all the time.

Host: Yeah, I mean connections and network are probably the final things that you can destroy because I think those are probably the only ones that you cannot wipe code yourself.

Guest: Or yet, God knows. But but I think the the general theme of you've got you've got to have a product that customers love and want uh because building the actual thing is not that difficult anymore.

Host: does it also mean that you have to like do a lot more integrations with a lot more products for a customer to adopt you?

Guest: Yes and no. I mean we there are few standard protocols that are used in the LMS world, it's LTI, the scorm, in the CMS world it's different and now MCP is coming to be. So uh not necessarily.

Host: Got it. Um I think that's pretty much what I uh planned for the conversation. We're almost at the end of the time. Uh what is the best way for folks to discover Uli? Uh where should

---