Identity as Infrastructure: Scaling Expertise via AI Clones

Machine-generated transcript from the episode video. Speaker labels are not included and some names and product terms may be transcribed phonetically.

[0:00] What happens when your clients can't get you level advice without booking your calendar? Not from generic AI, but from a clone trained on your thinking, your work, and your voice. These calls are starting to sound very monotonal when they meet after the agreement. Average 37 days. Why isn't there a AI native way for people to exchange information? Why do we even need these engineering leaders or whatever in these calls? Can they dump their brain? Clone is you can think of almost like a UI and the back end is data. The clothes are only as good as the data and the infrastructure. It's all about graph, right? Like we are behind the screens, we are building graph. All our customers have their name as the product. You are the product. You are the knowledge. We can't make or build wealth by selling that. You know, all these even knowledge professionals, they converted that into an asset. So our pro plan is $19 per month. We don't take anything. We're at the start, the creation of a new category. This is the way most if not all knowledge professionals will scale themselves. The last decade rewarded people who sold time. The next decade rewards people who ship knowledge as system. The winners won't book more calls, they'll scale themselves. Welcome to GTM Vault. Trusted by over 25,000

[1:12] founders and operators. Building the future of revenue. This week's guest is Vignesh Rabichandra, founder of MyOne, where he builds voice and text agents for experts who want to scale themselves without burning out. Before My Clone, he led critical database infrastructure at Cloudflare, traffic for more than 20% of the internet, co-founded Omnias, a Postgressbased startup built from open-source. So, Vignesh, I want to get right to it. What pulled you from databases into cloning expertise? So it's pretty much a journey. I was doing as you mentioned uh cloud web database then I had my own database startup in the postgress space. After that I left that I started like a two-sided marketplace connecting early stage founders like very similar ICP with engineering champions like a VP of engineers and CTOs like a classic two-sided marketplace or an expert network but then we did that after 6 months two things we noticed a lot of these calls are starting to sound very monotonous where the founder asks about the champion you know tell me what you do where do you work etc etc and then they get to the meat of it like tell me about your challenges and things challenge that we noticed is that you

[2:24] know there were a lot of reschedules like a founder and a engineering leader agrees to meet and then you know what when they meet after the agreement average 37 days in the meantime you know they do a couple of reschedules one of them don't show up you know something happens life happens that's when these two things led me like why isn't there a AI native way for people to exchange information can they dump their brain and then can anybody come and pick their brain or have a coffee chat or or even you know interesting deep conversation without them physically present. So that's really what led me down this rabbit hole of what we are building with and at what point did my clone go from experiment to inevitability? I mean we we did this around uh August end of August or I mean where we started doing deeply customer research then we started figuring out then who needs it so badly. It's not the VP of engineers and CTOs because well they have a day job. So what what's the incentive for them to go and create these clones and grind them? Um all those research went in around August and September. We just started talking to many people and see the love the pull from them. That's what we were like you know what wow this

[3:32] feels so different you know compared to all the products that we have built we were like no we never had these kind of customer conversations um like from coaches executive coaches consultants brand marketing consultant fashion consultant like now we no longer like you know restricted by any verticals uh amongst all these like close to 85 to 100 conversations we had with our ICP and we felt the pull and then we decided you know what let's just go on there. What similarities do you see between infrastructure and building AI clones? AI clones is the word clone is almost like a metaphor, right? Like it doesn't really, you know, end of the day it's all about infrastructure, all about data. The clones are only as good as the data and the infrastructure. So what that means is like having a very clean data set uh first of all imported having a good understanding of what the data is like it's all about graph right like we are behind the screens we are building graph let's say Rick okay Rick knows about say it's go to market like how how do we interconnect all of them and then build an hierarchy on top of that data or the graph so I would say I mean it's very much actually infrastructure play but morphed in the form of AI and clones why did you select knowledge professionals as your ICP well I mean they the huge body of work.

[4:43] It's it's as simple as that. Like all our professionals right like knowledge professionals when I mean these are like consultants, coaches, advisers, right? Like for them their knowledge is what you know really makes them stand out. That's why people reach out to them even software engineers or or anybody right for the matter of fact but these professionals have a good body of work in the form of let's say podcast courses substracts or newsletters I'm just thinking or white paper quiz assessment like the frameworks methodologies they have developed these over the last 20 years and so that's that makes it really easy for them you know to get started they can see the value immediately compare that to a service- based professional like a plumber or or you know or an electrician I'm pretty sure they we will get to them also or somebody will get to them too because everybody needs to scale for them it's a much bigger challenge like first they have to create these body of work to to ingest or provide this data set you wiped an early codebase and rebuilt it from scratch what went wrong there what changed and what finally clicked so I mean not like we rebuilt this this is the previous codebase so for rap I mean the idea that I just mentioned about the two-sided marketplace like an expert network connecting early stage founders and the engineering leaders.

[5:55] That's the codebase that I just threw. I was like, you know what, we can't we don't want to be maintaining two products or two projects. Even though that that was generating revenue, I felt, you know, it's just going to dilute our focus. We wanted to go all in on microphone. So that's one day I said, you know what, just drop the code base and then let's rebuild it from scratch. I want to talk a little bit about turning expertise into a product. Why is 15 plus years the break point for cloning? Uh it's just that our ICP like what we are just observing, right? like, okay, who's seeing more value? Who is more active? Okay, what's something common amongst them? Oh, okay. Look, all of these people have been in the industry for 15 plus years. Whether you are a a fashion consultant, whether you are a brand marketing consultant or you are an executive coach, we just asked them, so how long have you been doing this? They're like, hm, 22 years. H 18 years. Okay, 15 seems like a good number for us to, you know, map or build these attributes of our ICP. Can you dig a little deeper into I I'd like to better understand this concept of productizing expertise what is it what is mean for knowledge professionals as I said right like for them their knowledge is the product uh and a lot of them especially the SMBs their name is the product like it's it's a rigtm nothing else right

[7:08] like or or Qendrino's brand marketing ampa like all our customers have their name as the product I mean they even have five person working for them right like a CPA for example a small CPF firm that we onboarded but our her organization name is MECPA so you are the product you are the knowledge so where is the bottleneck in your business then how do we just scale you have to you know one go and hire more people excuse me or or train somebody to do that that's when we saw that around 243 2024 we have now this new capability in the form of technology where you can basically know replicate yourself in the form clothes personas the idea is that to exchange you right like this is not to replace anything it's just that how do we get more leverage as we all know that right like we can't make or build wealth by selling time nobody you know all these even knowledge professionals they converted that into an asset in the form of books in the form of podcast like anything or a group coaching so now what is that hey with with with your existing body of work right like why don't you build your own assistant like an AI assistant that can be you know maybe you educating your customers or um auditors like a CPA, certified public

[8:21] public accountant. They have to educate their market. All of them have to do some kind of education to their audience. Education is the first thing that they are offloading, right? They all their have their knowledge dumped it to us and they build their clone. Education is taken care. Second one obviously is to turn these education into a qualified business for them. Simple. Where where this education leads to? I'd been experimenting it with my own business and uh it fits very well into a nurturing uh workflow for sure. The c sorry pull the prospect down the funnel right while you're not there every positive engagement is um is helping increase the likelihood that that prospect converts into a customer. What I'm curious about is when evaluating the content that the knowledge professional uploads to train their AI, what piece of content works the best? Is it a podcast? Is it a talk, a doc? Or does format matter? Good question. To an extent, there is no such preference or anything. End of the day, you know, just so as a computer engineer, right? This is all ones and zeros. That's it. together you upload a PDF or a video audio they all get

[9:34] transcribed what does help is the having like a single person conversation right if it's in a podcast let's say there are five people now now we need to understand what Rick said in that conversation with that context and then add that aspect of his knowledge base so that just a direct section the concept that just makes it little bit challenge with the multi- people podcast and videos or anything obviously more defined is better like coaches for example they have like an assessment hey here is my 14 you question assessment that I do with all the initial prospects and can you now turn that into a into a persona that's very well defined if you think about it 14 questions here are the answers like a rubric um that has to be fit in there uh there is not so much wiggle room so much better use case for that role plays people are like you know what hey let me do a role play I have been doing helping with uh people finding jobs that's one of the things that I do or having like you know tough conversations with their bosses and peers okay I'm doing that now it's kind of similar set up. They have a methodology that we are now able to replicate that much better.

[10:34] Another thing that I've been, you know, it's been a few weeks that I've been using my clone and um one of the things I'm debating is at what point do I offer a premium tier and at what and how much of it do I keep as a premium tier? Can you walk me through your perspective on what becomes premium, what lives inside the clone, what and what stays free? So in our like the free plan right still the pay it's a payment processor obviously we use stripe so that's that's how we get you know payments through the through the AI personas so for context what trick is asking is like hey can I monetize my knowledge right like why am I building putting all this like you know form of courses and podcast so at some point in time we have to turn that into revenue which is very well what we would like our users to do so uh nothing wrong so anything other than our free plan right users get to keep 100%age of the revenue their clones generate we don't take anything. It's it's the user's knowledge. It's users visitors.

[11:31] It's users revenue. As simple as that. Rather, we charge by like asking a monthly subscription. So, our pro plan is $19 per month. One night, that's it. So, you have a $19 um you know, agent running for you. You charge maybe, you know, 20 bucks. That 40 bucks is is the users. They get to keep the entire $40. We don't take anything. Okay. And just to double down, a lot of our professionals also don't even charge them because they are adding it as part of their retainer service or they are providing it as part of their premium service. They can also put authentication wall in front of it, right? Like there are two walls you can put. One is the authentication wall, other one is the pay wall. Let's say you have like a publicly available like a clone or persona that anybody can come and talk to it or pick your brain or you know do a role play. Okay, sure you can put like some pay walls to make sure that you know you are offsetting your your knowledge cost and the infrastructure cost. Then people are like you know what I don't need to charge them because I have already charged my clients this is like a premium service or add-on service that I'm providing. Sure in those case they're putting an authentication that's about it. So just want to you know um clarify that authentication payment is through directly through clones is not the only way to monetize the their monitor.

[12:41] So I want to now get a little more technical when I talk to a clone what's happening under the hood. So as soon as anybody we call them as a visitor right like a visitor is engaging with the clone what we do is just to pull up the the persona that the clone that they are talking to with the visitor with all their existing knowledge base you know say I'm I'm for example chatting with Rick's uh persona by the way I mean these words personas and clones are interchangeable right like so in in my clone we have like so many personas a user can create so we start up that persona with the objective specific objective let's say I'm talking to Rick's roleplay persona that knows that hey Rick is 30 years of you know doing go to market I've done this and here is all Rick's knowledge for this persona and here is just meant for role plays like a early stage founder who is let's say trying to talk to the prospects how how should be this 20 minutes conversation should go all this knowledge gets all added as part of the system prompts to the models so we use web RTC behind the screens that's the one that's providing you know these connection links between the visitor and the user uh with this system prompts

[13:50] that I mentioned and after that uh you obviously use a rag hierarchical graph rag is where it's pulling the knowledge from let's say the user says the founder says that you know what I'm building for security so now it needs to go to rig's knowledge base and find everything related security related stuff not database related ones why do most rag systems fail in production okay uh why do they fail it's it's going back to the very first question we discussed right like clones and infrastructure it's very clones and you know database infrastructure it's it's all about data at the end of the day garbage in is equal to garbage out so a lot of the rack system that I've seen is not have like a clean data source at the first place to begin with duct tape nothing works it's kind of like okay demo works but not like a production grade system um lot of these parsers in general like parsing is is very fragile idea right like the idea of parsing in computer systems where you have some data and then you parse that it's just like a bunch of rules is how the data is parsed from one format to another format and that format that parser is always tend to fail. So there are some inherent

[14:58] challenges in the in the computer science itself. I mean engineering itself how these processes are built but obviously you can build very reliable system right like you can get to 99.999% let's say you know you're you're nailing like write a lot of good testing end to end testing unit testing so to make it even better and robust. I'm sure you've made some mistakes in your own pipeline. Could you talk about how how some of those and how did you fix them? So all the time if if at all we get it right that's when I'll be surprised right like how is it working properly one we recently talked about is the is the latency is one of the things that we always obsessed about is that reducing the latency right like who don't want their other person to respond as soon as possible right like whether it's an agent or a human we all want immediate response so on that note like we were using a different embedding models like open ones bigger models bigger dimensions uh they were taking a lot longer than than the other models mod that we tested. So we felt like we can shave almost like 25%age in the latency by just changing the embedding model without losing the accuracy like without losing the quality. That's a very big you know clause. It's not like just changing just not reducing the latency but keeping the same quality.

[16:10] Speaking about accuracy and latency, how do you balance these and consistency and realism? It's a trade-off like you know you can give answer faster but what if it's just like crap answer compared to like I'm going to give you the best answer but you know come back tomorrow both of them is not a good answer not a good position um it's delicate but we kind of starting to understand um the nuances now what's a good enough answer uh without losing the user or the visitor definitely latency is one of the biggest factor I mean even at the at the you know in in a terms you might even compromise a little bit on quality, but latency has to be really really good. Another thing I've been thinking about is like what matters most? Is it sounding human, being correct, or speed, being fast? Wow, amazing question. I have to just put a vin diagram like all three and then put the center. That's that's the right answer. There is no such thing as, you know, you can just speak super fast, but if it doesn't sound human, well, that's not really going to cut it. At least the promise of the iPhones. That's that becomes just a, you know, a chatbot. Also, you know, you can't just take forever, right? Like saying that, I'm going to really sound like human, but give me 5 minutes.

[17:22] That's just too long. So, we're at the start, I want to say, of the creation of a new category. It's obvious that this is the way most, if not all, knowledge professionals will scale themselves out in the coming years. I'd like to learn a little bit more about your GTM strategy for this new category. Amazing. Love it. I've been doing a lot of GTM these days. So, so my team members mentioned me know you stop writing code. That's the message that I got from them. Yes, I did spend a quite a bit of time writing code. So, but going back okay, what's the GTM strategy? First of all, you know, thanks that you are, you know, folks like yourself seeing where this is go going, right? I mean, that's fine, right? But the point is that you need to see where this is leading to, right? That's much bigger challenge, right? Like imagine that I have this knowledge. How else I can scale this? have to train a human like an assistant in turn or I have to train a machine that can speak in different languages that can be available 24 + 7 there is like all the time you know there is no rest for it all trade-offs combined my bet is that you know what you much better train a you know computer or system than human I mean nothing wrong right you still have

[18:30] a place for them you know we will all will have a place that's not what I'm advocating but a lot of these like coffee calls qualification calls knowledge education I'm I'm very bullish that it will be you know transferred. Okay. Go to market systems uh very near and dear. It's really without a doubt that a knowledge professional is capped by the number of hours they have right within a given day the amount of hours they can sell. So um this really enables the individual to scale themselves out. Absolutely. We are also seeing some service professionals is the next in line who's who is coming up right like even as I said some examples like a realtor or a CPA um I mean realtor is a very good example right because they still need to do service like go and show the house and you know what not but they still also need to do some amount of education to their to their audience so we are seeing the next set like how you know we can expand our time like we just picked the knowledge professionals because it's so perfect for them okay once we done that learn that who's next in the Okay. Okay. Service professionals and these are like pretty pretty big industry right you know knowledge professionals so many um attorneys you haven't even talked about or touched

[19:41] upon but okay going back still to your GTM question right like so we are going up in the PLG motion not really you know like a one to five person users can come in like yourself who sign up for free we give a generous free tier so they can actually play around and really get an idea that you know they get to see the product in action and once they are like you know what this is Okay, that's the only time they, you know, go to upgrade to a professional plan. Obviously, they can try the business and, you know, finally enterprise. The motion that I clearly see is that it's going to be a lot of partnerled, channelsled. It's not going to be just us trying to sell to every single one of them. For example, a lot of our auditors whom we work with, they to be don't honestly don't want to be setting these up, right? Like I don't want to be touching computer or whatnot. You just get the job done for me. So that means typically somebody intermediate who's helping do the job. I mean we also can't be like you know go on every one of CPAs and set these up and then we just become like you know cognizant or an accent at that point like what problem is that solving?

[20:43] Yeah it's it's VA cost right that's the one that we are replacing today when we tell them that you know if you have like a virtual assistant of some sort what is the what do they do? They maintain your calendar. They help you find you know have these qualification calls and then they bring you to the table. That's that's a cost that's just saved. So all things we can do right like one is the time saved for an individual we are saving the time or if they have a virtual assistant or some kind of an assistant associate that's a that's a that's a time saved. Obviously we can also equate to cost generated revenue generated. We just talked about that you know users are putting their stripe wall stripe authentication in front of it and start generating it. The third one is that they have to basically like build a legacy right like they are leaving their impression whatnot. These are at least how we noticed that the people are you know why they spend some money in any of these products or services. Time saved revenue generated or leaving a legacy or are leaving you know good impression?

[21:36] You mentioned that you're going to market with a subscription model and a premium tier. Um do you think this might evolve into perhaps more of a usage based model down the line as you know the company matures and absolutely um gains more adoption? Yes, it is usage based because if you don't if if our users don't use us, they don't need to pay us, right? We don't need to charge them. We know we can pass on everything that they save and if our users use a lot more of our agents, then obviously we need to charge them more. So very much into the usage based model. Going back to the GTM, I want to touch upon two things. The motion that we see is very close to have you heard about go high level G what is a it's a CRM like very much used in the SMB space. A lot of agencies whatn not they all use this go high level like a CRM uh that's one the other one is clay yeah I'm familiar with clay obviously so that's the same model in a sense say you know we have to educate our partners and and you know channels they are the ones who are going to likely set this uh clones up obviously any user can come like yourself come and set it up but if you are talking about SMBs like 10 to 15% like a CPA firms or a legal firm

[22:46] then we need someone in between speaking of channels. Where is demand actually coming from today? Mostly it's from LinkedIn and Reddit. Those are the two best performing channels. Yeah. Can you speak to me about how you're creating demand through LinkedIn and Reddit? Uh obviously the big piece is outbound especially in LinkedIn like we we know our ICP then we hit them up uh with a personalized demo. Like we create these one minute personalized demo with a Loom video um of mini role playing with their agent. So if they see it they're like hm okay I immediately see the value now I have to decide whether I it's useful to me or not okay that at that point you are making a decision very good scalable but still you know we have automated so much of it so we can still cranking out a lot of personalized demos but other other than that the materials right like a lot of founder sales I have been you know constantly being out there and and you know trying to educate the market basically it's not about you know replacing you it's about giving you more power to you it's like almost an extension of you Other message is that you know listen your users and prospects and clients are already talking to AI.

[23:52] The question is like you know are they know do you want them to be talking to your AI or a generic AI. So those are at least you know the messages that we have to spread. Reddit is a lot more of blogs and technical details like for example all the technical things that we talked about we go and share them right because Reddit has you know it's pretty strict when it comes to how much of you know promotions that you can do. So, but even then it's fine. Even if you just give like a goodwill, right? I like let them come and see what we do, what we have to say and I'm pretty sure hopefully you know that they will all turn it into some kind of karma and then eventually turn it into users. Got it. As an transforming from an infra leader to a founder, what leadership lessons transferred directly? Oh man. Uh I'm still learning by the way. I don't think I I wish I I bring a lot more. And it's also a good question to my team now that you asked me. I have to go and ask them that you know how do you guys see myself actually uh being a leader like what qualities you like and don't like but see it's maybe communication you know one of the things that I really stressed or like you know give more weightage is is communication clear communication nonviolent communication is super super important

[25:01] other one is like uh we I mean at least I operated this principle called give and take it's it's basically a principle that Adam Grant a professor from University of Pennsylvania who wrote a book on this topic he said how you know givers eventually win. So I read that when I was you know working full-time job. So I picked that up using the same oategy even when I'm builder. It's like give first you know that's why we have like free users premium you know give enough then they will come then they will come. Absolutely. I love that. I love that approach. Definitely you give more than you take. It needs a long long longterm thinking. Some of these are sometimes you don't also it's it's a very tough world especially current world is like everyone is stressed you know what not even if you give doesn't mean there is no guarantee that you'll get back but even then I still operate in that same mindset but we'll see yeah that's that's one of the issues of being a CEO right how do you balance short-term results with long-term patience and and thinking how did your hiring lens change how did my hiring lens change to an extent hiring lens didn't change much you know Still even when I was working at Cloudflare um I used to cold outbound to candidates like I write personal emails saying that I've gone through your GitHub I've seen your contributions

[26:11] do you want to do the same thing full-time definitely will pay you beat the market what not do you want to join that was my strategy of hiring and even now I'm doing still the same obviously we don't have the brand something like cloudflare so it's like what's you're a 5% startup okay but still the process remains the same it's like hey find the really top candidates whom you think would be a good fit and then and then just reach out to them. I don't like doing any of these, you know, structured interviews at all. At least for now, you know, it's it's working out well for us. We go and find those really uh what do you say like rough around the edges candidates and then bring them in and then and then you know try to fit them in. How does a community influence product design? I mean it's a big part as you can see right like we are not talking about enterprises. So that means we we are working directly with the individuals they are very vocal right like I get a lot of bug reports from WhatsApp cold DMs like LinkedIn whatn not it's just going to be huge part any of the recent EI companies that I'm seeing now like from the ground up or just to say yes it is going to be a big part of uh my cloud are LLM's competition or leverage LLM's competition or leverage wow Rick this is this is where we are getting the the juicy part I still think it's leverage right like you know yeah you know you

[27:23] have you You can use this somewhat kind of a smart box to your to your benefit, right? Like as long as you can. So as like anything that we have seen any other cycles like mobile or internet or any other waves. This is just going to be like you know additional productivity make our lives better. It's still going to have its place. It's just not going to be like you know what I'm going to replace everything. I I really don't think so. After all it's trained on publicly available knowledge for the most part, right? It's not generating anything new. I I really don't believe that you know that it's just going to like come up with groundbreaking something on its own. I mean there is a lot of talk around that like they can generate what not. It's still like you know LLMs are just spitting out the next character that's all. So you know we are not going anywhere they are going to be around. Uh it's just we need to figure out how both of us exist. LLMs and humans.

[28:12] As an early adopter of AI nowadays, it's so interwined or it's so integrated into my workflow with nearly everything that I do. I I can't imagine how someone who hasn't started to integrate in their workflow could compete in in speed, accuracy. It's just changing so fast. And I want to ask you, what does AI native mean in 2026? Great question. So what does AN80 means in 2020? We can talk about you know it depends for whom right like we can talk about for an enterprise or we can talk about uh our ICP. So let me just stick to our ICP because that's much more relevant for them. It means adopting a like what you just described in every part of your workflow you know end of the day it's about building your obviously the pipeline right like for any of the knowledge professionals or any individuals SMBs whatnot. Um second one I said um converting them right into pipelines and then obviously retain them and then they give you referral back.

[29:13] This all this pipeline how can you use CI like how can you leverage AI for example meetings taking and I use uh email summarizing or document generation or uh creating all this PDFs invoices like legal contracts. This is all part of what we do day in and day out. If you can use AI everywhere to get more out of us then we become native. The only big thing you know that what we are propo proposing I mean at least with micron is that build your own clone your own AI right like don't just use the the big ones obviously use those but they don't have all the workflows they don't know what it takes for it it's very big broad these larger LLMs now you need something that that is much more you know attractable malleable for you basically translate that into your business workflows finding qualified leads or converting them that's what we are trying to do Our biggest competition is that generic GPTs in charge GPT you can create something like custom GPT that's just like a bunch of basic prompts that you can put together that does the job and we are saying that we are 10x better than the custom GPT so for example our users use it to qualify sales leads they have this meeting conversation as soon

[30:23] as they finish finish the meeting they dump that to my clone and my clone gives them back that you know how the conversation went like a rubric that they have and our users try to do this using custom GPT and it fails every single time. Yeah, I can imagine it's be too fragmented well when working with custom GPTs whereas you're providing a end to end solution correct yeah end to end solution like with your data now it's all backed up you can always you know go back to your data and then you also have your framework and workflows integrate all of that now it just becomes a much more robust uh like you know workflow that I can use day in and day out and where should ethics draw the line here where should I think I We have you know definitely made it very clear that it's only used know at least we proponent using it for the positive way and we going to be a police here right like making sure that any kind of non-ethical activity we don't allow in the platform like trying to use other other users you have like a strict guidelines data policy that you we don't allow any priv privacy violation and things like that it's going to be fun we have to chase as like any other technology right like uh

[31:32] 99%age use it for the good purpose with the good mindset And then there is always one percentage which they're trying to find shortcuts and our job as a platform provider is to make sure that we curtain that one percentage and we will do clones become products, media, infrastructure, businesses. Wow. I mean to an extent it can be all of them. It can be a media. Sure, why not? Like I I mean I have your clone also. So now we I could have even done this conversation with your clone then that this becomes media product. Yeah. If you can put up a wall in front of it like do the role plays you know pick my brain or I'm going to analyze your sales contract that's your product influencers why not now you're going to be like you know let's say I'm going to engage with your clone saying asking I'm in the market for a sales cold email campaign which products you recommend you're going to promote let's say instantly well that's your influencer marketing or affiliate marketing so I mean I just see sky is the the limit for where we are so many use cases and what happens to consulting in a clone powered world.

[32:33] Yeah. I mean, so so what we say is that you know it's still going to be there, right? Like you still want the human judgment, human connection. Like we have like consultants who have 20 plus years of experience. They are adopting the the idea for them is so no-brainer. It's like a best of both worlds. I'm still going to be me. You know, you you can't represent my know you can't replace my connections, my experience, hard lived experience, my judgment. But you can now also get the the kind of chat GPT experience. So that's why we say, you know, it's a best of both worlds. I definitely believe that we are just going to, you know, coexist and what we see as these clones and personas as an extension to scale them. I'd like to move on to the rapid fire section of the pod. One myth experts must delete. your years of experience doesn't correlate doesn't matter who cares habit to adopt um voice based voice dictation one GTM metric that matters retention one expensive founder mistake not paying enough attention to distribution the last decade rewarded people who sold time the next decade rewards people who ship knowledge as systems the winners

[33:45] won't book more calls won't hustle harder won't scale headcount, they'll scale themselves. Vignes, thanks for showing us how identity becomes infrastructure. If this episode changed how you think about leverage, share it with one person. Still trading hours for income. This is GTM Vault. Build systems, not stress. Tech founders and VCs careers lessons. GTM