How This GenAI Startup Got VCs Interested Without Selling The Product

People always say starting a startup easy but raising funding is hard. What if I tell you it's wrong? It's opposite. If you're able to create the best product and VC is going to stick the money to your hands. That's exactly what happened to Locus AI and within like two months of creating the product even before they launch a product they have two VCs directly invite them and saying that hey want to talk to you want to give money to you. In today's product insider podcast, I'm going to dive deep regarding how exactly is able to launch such a successful product with early adopters and everybody's raging about it and using [music] AI agent to organize your knowledge database. Hey guys, this is Dr. Nancy Lee, a direct product feature in Forbes. I've helped thousands of people land a dream PM job offer in fan companies and unicorn startup and continue [music] to get promoted as a product leader. In this channel, we have a tech trends and free product management training. Like and subscribe and turn the bell button so every Tuesday you'll be notified with our [music] new video. Hi, welcome Locus AI team. Welcome to product insider podcast. How are you guys doing? >> Doing well. Well, thanks a lot for having me and my team Locus A over here. >> Beautiful. Hi Shbam. I'm so glad to have you here. I know you are the CEO of Locus AI and you guys actually created amazing AI agent that is able to organize all the knowledge database and I'm actually the very first early adopter. Really want you guys to make decision for me using a agent. It is totally groundbreaking. That's why I'm very excited have you guys to share all your amazing knowledge developing your AI product and your your uh the knowledge and AI agent with us together. So uh why don't we do this? Um Shbang you have your whole team here. Why don't we do a quick introduction? What's your background and how you actually I know Shbang you transition from traditional product management into AI product management and AI entrepreneur. Can you tell us more regarding your background? Let's get started. You first Jabbang. >> Absolutely. First of all, thanks a lot Dr. Nancy Lee for inviting us on your podcast. Uh talking about my background, I have over you know seven plus years of experience in the industry. uh working on the retail, e-commerce, banking and now working in a healthcare segment. Coming from the traditional product management background, you know, I do I do have a background in the development as well. I worked as a developer then took the product management and eventually with the PM accelerator which helped me a lot in transition to the AIPM you know with a focus on developing one of the best tool which is locus AI and we are fortunate enough to stoop second in the entire cohort but a quick kudos to you know everyone to the locus AI team >> beautiful congratulations and yes very excited about your fast growths and especially your team actually is the smallest team but took the second place and it's I just very excited about your product. I think people everybody watching a demo was like yes let's all like jump on the product. Shabban before we dive into all the excitement all the locust AI literally people already start using it and you guys also talk to like investors the two weeks before you actually finish AI boot camp. So before we get all the excitement let's understand who is the brain power behind the locus AI. So let's introduce your team. >> I am Abas Rahman. I am a backend engineer at Locus AI. Hey everyone, my name is Jun Cho. Um, I am the UI UX designer for Lotus AI. >> Hi everyone, I'm Sadira. I'm working as an AI engineer for Locus AI. >> Hey everyone, my name is Taran Malipati and I'm working as a data scientist for Locus AI. >> So very excited to have you guys here. So let's dive deeper a little bit more regarding Locus AI product. Um, Shbam can you tell us more regarding what is Locus AI, why do we need it, what the benefit using locus AI and maybe show us what's look like. Absolutely. That's a good question. Locus AI is basically a memory layer for people and AI agents. Locus AI is basically offering a unified platform for integrating different communication platforms like Gmail, Slack, notion, Java conferencing, the list goes on. The core pain point what we identified is nowadays we have a lot of tons of data silos you know uh located everywhere. Whether we talk about Gmail, Google Drive or you know some of the other communication tools, we do not have any common unified platform where we can just put a keyword and retrieve the entire information. As per the Harvard Business Review, I found that it it almost you know takes like 5 to 7 hours of every person you know like every individual spends more than 5 hours a week in just retrieving the data. So we found you know let's start with this pain point and we conducted over 30 plus customer interviews and we found that out of 30 24 people were willingly you know to invest in such product. So we thought okay we got our idea validated. So idea validation is done. Now let's talk about the technical architecture how it should look like then we start building the you know the low design high level design and or the system design of the locus AI I really resonate regarding people spending like hours searching for file was very true especially the file from years ago is very hard to find especially we have so many different tools and for example we use Google doc we also have hopspot and we also have discord and a variety different things I believe one of the features you guys have is actually helping me to make decisions. Let's say people message like hundreds messages on Slack every day and you're able to your AI agent is able to help me to make decisions and flag those information on Slack. Hey, this is very important this your priorities. Can you tell us more regarding that kind of like stress relief feature to your AI agent? >> Yeah, of course. So building Locus AI we I was fortunate enough to have you know a good team of people who have remarkable experience and have producted to the best of their ability right and talking about the locus AI features we have like core features which are do you want me to demonstrate a quick demo well will that be okay so this is how the UI looks like this is the landing page of Lucas AI which basically tells you you know why Lucas AI how it works and we also have provided a user guide for the new So let's start with the login page. It has a terms and conditions. Let me agree. Click on get email. So there you go. As soon as you click on the dashboard, this is how the dashboard looks like. The dashboard basically has a search bar. Let's talk about give me the latest information. Let's see what it gives. It has a back end which is basically a cloud API for structuring the data. Right? So as soon as I click on the asking it will give me a lot of information. Right? So as soon as you click on the information right it will give you a very consolidated summary participant source status in conversation and even the biggest USB of the locus here. If you click on the original, it will take you exactly where the message actually resides in the conversation. You can see this right. It saves a lot of time. Similarly, in the memory layer, we have a decision action item in blocker. If you click on anything, right, it will give you a very consolidated summary along with the data source connected with it. Right? So this is something core and a team post you have a calendar filter where you can actually put you know from which date you want information. So for example I choose this and you get some information for the week. This is how you know the locus I works here you can see the number of services which are connected. We are you know wanting to know more features if user wants full context or just want the core knowledge. We do have the privacy parts as well. The biggest concern what if it can retrieve the grinding information. So we have defined our guards very strongly to make sure that the data is secure since the data governance you know plays a very crucial role when it comes to the data retrieval part. And we also have a chatbot and a look. It can only answer about the local city. >> That's a lot of feature. I like that you putting guard rails and because this is actually one of my questions regarding all the information, right? There's a lots of healthcare information, banking information, everything, right? So you guys are re already pulling the guard rails. So have you test this out with user? Do they fully trust AI manage everything about their information? What does the user say about it? So yeah uh uh to answer your question we do have some users you know who are using locus AI for in the pilot phase and they told me that you know they really like the product how the product basically offers the unified platform or organizing the all the information from different sources but at the same point of time they do have some concerns when when the data privacy comes in right we are currently working on defining our guards making sure we are completely compliant and for the phase two whether we comes to the healthcare banking uh you know information we would be we we be we will be considering you know the HIPPA GDPR and other regulations to make sure we have our due diligence. >> Beautiful. I like how you guys building a product in term of different phases and we start using the normal information knowledge regarding what's available with the day-to-day work for example all my work information I'm happy to have your AI to start manage for me and and then of course for all the advanced healthcare hippa takes long time so that's why you get to the different faces approach this is very smart now Shioban let me ask you this question actually having this AI agent running in the background organize all the knowledge database is actually I believe very challenging. Uh and then but you guys actually finished ripping this product within just two months. It also like two weeks earlier than even before we had a demo. You also start like talking to like investors two weeks ago and this is like really fast. So what's the secret developing a real life AI agent product in just two months? Um what's your lesson learned from building it so fast from zero to one? So to answer your question, I wish there was a secret recipe for building an AI product, but I was fortunate enough to have a good team of people who work day and night in order to turn this entire vision into reality. Uh so talking about the the integration part, right, it comes with a lot of challenges when you are connecting different data sources. Before MVP1 of Locus AI, we are using Slack, Gmail, and notion. So for Slack and notion, we were pretty much, you know, fine when it comes to the integration part. Gmail of course has a lot of issues because if you see in the Gmail right it has inbox outbox family and spam a lot of filters over there. Of course it it took a lot of challenge. It took some time for us to resolve the issues of integration the Gmail and the locus AI but of course we were fortunate enough to you know integrate those things uh in the locus AI that is fully live users are using it we are getting consistent feedback on a daily basis but I would like Sadira to throw some light on the integration challenges that we faced so far please. Uh yeah sure. So uh to start with answering how we were able to finish it within uh two months to get the MVP. Um so we kind of uh dissected the entire tasks into smaller problems which we could target each week and uh that's how we were able to organize our tasks uh with we we followed a system in notion as well where we distributed the task within the developers weekly and were able to meet the deadlines. So uh with issues regarding integration we had to connect these enterprise tools and uh different data sources to our agentic multi- aent AI system so that these agents could securely access and act on that information through these tool interfaces. So uh we did face a little bit of challenges while we were connecting Gmail like uh Shubam mentioned but uh we were able to uh sit through uh um each challenges collaborate with each other um effectively and uh we solved the issues here >> and also Dr. to add in uh you know we we were fortunate enough to finish this before the tagline right and and the second part is the talking about the VC's part right so u I got like couple of VCs you know reaching out to me on the LinkedIn about the locus AI we are still you know consuming them and still in the talking terms how we can potentially you know launch a locus AI at an enterprise level so right now we are still relegating our idea in terms of users willingness to buy to pay a certain price right for the locos AI. So we would be launching on a premium uh model of course in the coming days and we are already working on MVP2 to integrate VA conference and discord as well. >> This is beautiful. Hold on. Hold on. I think this important part here VC reach out to you on LinkedIn. You didn't pitch to them. They found you. Reach out to you on LinkedIn. This is like how crazy that. Tell us more about this. >> Yeah. So I was fortunate enough to you know getting VCs over the link and is like the biggest [laughter] miracle but of course we are still validating our idea. We are I'm preparing you know the different pitch decks we have applied to some of the combinators. We are still preparing for Y combinator for the upcoming eggmit and making sure that we have the real users to work right I mean we are still in the process of finding a C comp as well the locus AI and we are hunting down for the co-founder particularly with sales marketing and go to market. So this is where we are currently standing. >> This is very exciting. Hold on. So basically your product is so popular with those early adopters. So people are raving about it and you guys also like share your takeaway about Locus AI and then VC basically saw it and reach out to you guys. This is like mind blown mind blown. This is like next level of success like take my money tell me more. I'm interested. I I can see this actually in our last episode we invited Abham and he actually started his own AI company and a year ago and now he has 15 employees working for him. He just signed another million dollars contract with another 20 billion dollar clients. It's like crazy. He is facing the same thing. Even the moment he finished AIPM boot camp, he has a four like architecture design company reach out to him saying that I heard you have this, you have this. Can you show me a demo? Can you see how exactly you can solve the problem? So, it looks like they have real pain, right? They have income is a is a great indication that will let to one year later he has a 15 people working has a real like companies and like signing like million dollars deal right now and I see you guys are on the same track. Literally the VCs and the people consumer basically like we heard about this, WE WANT IT, PLEASE DO IT. This is like mind-b blown very exciting. So proud of you guys. >> Well, thanks a lot. We are still in the process you know reaching out to multiple VCs at least we can secure a deal as soon as possible but I'm fortunate enough to have like a team of seven people as of now working towards the MVP2 and a lot of MVPs in the coming days of course >> so as of now our next release is targeting in the middle of September so yeah >> this is beautiful beautiful so I want everybody who listen to the podcast right now because building a product it's just beginning and lots of people say oh I'm landing a job but you might be the one who can create job for other people which also lead to a very important thing in the step one for people step into the age of AI is actually start developing and gaining hands-on experience. You may go on entrepreneurial track have we hunting you down give you the money take my money or you could also be using the experience to land AI parlary job or AI engineering jobs. So number most important thing is you must start developing and launching real life AI product with real users. So, a free checklist, a cheat sheet of download the top 20 AI product ideas to gain real life AI experience. Um, please go to pms.io product ideas so you can download and start implementing right away. I'm also going to uh link it in the description of the show note. Now, Siobhan, let me also ask you other like technical questions regarding what tools and model have you leveraged create your AI product right now. I assume you leverage all the best models and tools. So what are any technical uh tools and models you recommend people to check it out? >> Well talking about the data AI models. So to structure the entire locus AI we have three connectors at the top which sits at the foundation layer. For the Gmail we have pubs sub service of Google for Gmail for Slack and notion we have the polling mechanism and the web hook which is using and then coming down deep diving into the uh you know the login part GCP is one of the key ingredient for it and of course the AI models right we have the entire the core functionality is the semantic analysis right because the way locus AI retrieves data it basically has a semantic layer which reads the context, categorizes them, what is the decision, what is the pending item, are we able to identify the blocker as well. So for that we are using a cloud API, you know, particularly we are using the model haiku. It it is pretty old model as of now. Currently the opus is already there. Opus 5 is the latest one. So we are using cloud haiku at the back end because of the fact that the cloud is a pre-trained model, right? it automatically identifies what is the decision you know and all all other items. So we are basically using that API to structure the data to extract the data from different resources >> and this is how the entire web pipeline is having right so the vector embeddings is done by voice cloud API is taking care of the structure of data and extracting sources from JS like and notion and basically we have a context graph and the knowledge graph right which is basically setting the entire web pipeline and this is how the data injection to you know the entire querying part and basically what you see on the dashboard of Lucas AI. >> This is so impressive. I love how comprehensive and different tools and messages and and models you have you guys that use. What I found out lots of time is that um the simplicity of the product make it easy for people to use but to enable the simplicity we require very strong technical on back end to design in a uh in a very strong way. Uh this is so exciting. Uh all the different kind of like AI tools and models and we're going to also link in the description the show notes so people can use at a reference to design your own AI product as well. And now let me also ask this question and let's I think everybody who listen to this right now they're so excited they want to use your Locus AI. So where can they sign up right now? >> Yeah to sign up basically uh you can go to the locusaiapp.com or registering yourself to use the Locus AI. >> Beautiful. I'm going to link it in the description of the show notes so people can start using and sign up and experience this yourself. This is a pure transformation. So let me ask you guys this question today. We have designers and product AI prime manager AI designer AI engineer here on the call. What I uh want to ask you guys what advice do you have for other AI entrepreneur AIPM AI designers for them to jumpstart the career in AI? Um Shbam why why don't you get started? So of course uh that's a good question. So first of all have all the foundation skills which are pretty important right. So coming from a product management background I do have a developer background as well. I combine my both the technical aspect and the business aspect right to make a technofunctional position right and at the same point of time leveraging your AI capabilities you know learning different tools how integration works and all that. It basically helps you to you know give a new mindset in terms of to think about outside the space right basically what we talk about the out of box capability to build an AI product so it it's just about you know keep refining yourself learning new skill set build real products you know for your portfolio which can actually help you in getting somewhere in the entrepreneurship or you know even securing a good job. This is a great advice and especially for people who's interested in gaining hands-on experience by building a team of developers and designers and you can directly go to our website pms.ioipm to learn more. Uh you can also directly book a free career consultation call with our team to see if it qualify for the program. I'm also going to link in the description of the show note. So let's think about this. We also have a engineer in the room right now. So uh Shadier, what advice do you have for other AI engineers nowadays? Uh yes. Uh so one advice I would give is to so right now AI is developing so rapidly and there is new technologies out there every day. So I would say to keep yourself up to date with all the tech stack and the knowledge that is being put out there so that you can be wise and uh leverage what is necessary to be able to use at that point of time in that problem. Yeah. >> Beautiful. Very well said. What about data scientists? Terrone, you are data scientist working on locus AI. You must have lots of advice to share with other data scientists and how this is going to change their career. >> So as a data scientist, I would suggest that if you're like starting as a data scientist, I would say start slow because data science is a long journey. So start slow, start small and eventually like progressively consistency you can eventually build up. Don't take up like large data sets and start like filter them immediately in the beginning. It it'll be confusing. That's what I would suggest. >> Beautiful. Do you think AI is going to replace the job of data scientist? >> I actually think it's like the other way around like because data science is like helping AI. So if anything it will be more beneficial for us. That's what I personally believe. >> That's true. That's true because like garbage in garbage out a data is a field of AI. If the data quality wasn't there your AI model is not going to turn out good as well. So basically there more demand for data scientists because of AI. That's very very insightful. Uh what about designer June? So you're AI designer. U what advice do you have as a designer out there and do you think AI will replace designers in the future? >> Um yeah this a this is a problem that people always discuss among the designers community. Um based on my own experience of designing for Lotus AI I think um AI first they are like very powerful to help designers uh validate and refine their ideas in the process. uh which is something that I think at current stage AI is still um could not replace human being in that perspective. So if you're a designer I would advise you to focus on uh enhancing your empathic design skills um to learn more about your users and to just refine your craft so you could uh guide AI to do that p so perfect work based on your own aesthetics. I am always the the big believer of growing and growth mindset and I believe life is a school, life is a university. We should always upscale ourself and prepare ourselves for the future and I want to ask all the guests today how do you grow yourself professionally, personally, what resources you recommend for other people to grow themselves. So let's do this uh like quick tips right here. Siobhan, what what you have? Well, I definitely uh you know uh recommend people to you know be consistent in terms of the learning curve, right? Because it is very important to have a foundation of skills and keep building your AI skills as rapidly you know the AI is growing pretty fast right and at the same point of time invest your time in building real products right reading newsletters product like a team expert is one of the best place to you know have all the consolidating information and to grow yourself right the effective collaboration and all it >> beautiful so Shia how do you grow yourself. >> I would say that we have all the information in hand with us. We have everything on the web. So you just have to go look up YouTube videos like how Shubam already mentioned newsletters and documentation. So just you you you'll have to keep yourself moving. So that's about that that's what I would suggest. >> Totally. And everybody we have uh by now today I just got notification from YouTube. We post over 1,000 different videos on YouTube. At the celebration, we have lots of free resources for people to learn and many of them actually we have AI engineers literally join Amazon as a AI engineer job and also of course thousands PMs join Google M Amazon with the AIPM or start their own company very exciting all starting for free on our YouTube channel. You can just search Dr. NV director product learning for free on YouTube and LinkedIn Tik Tok and other resources as well. So terum what advice do you have and how would you grow yourself and and resource you use to grow yourself? >> So I would suggest that so when the AI wave was coming in people who are aware of it could like tackle and stay but people who are not aware of the AI wave they they struggled and they had to like again like fit into AI. So similarly I would say like stay up to date and uh at least once a month like look at the articles and stuff of what's the upcoming tech technology and models that are coming up and sometimes like play around on with it maybe like cloud or new cloud models and everything so that you're aware and uh you don't get lost in the AI wave or like any other way that's coming in the future. >> Beautiful. This is you can tell the advice coming from a data scientist as you directly read and study those AI models and things are getting technical nowadays. Uh what about designers on June? How do you grow yourself a designer? What resources do you use? >> Um, as a designer, I found there are um actually a lot of good resources among the designers community if you dive deep. Um, so there are a couple of people who currently built Loro designer meetups and local workshops. Um, I really recommend people to dive into uh actually LinkedIn to see what are uh those active designer communities in your local city. And um I I actually get a lot of useful insight and resources by reaching out to those people who like to share their experience and host events. >> Oh, beautiful. That's the power of networking and get your access and and into the space of AI. Very useful. Thank you so much. So now my final question, if anybody want to get in touch with you guys, where do they go? Do they go to your LinkedIn? Tell us more. Of course, if people want to contact us, you know, they can reach out to my LinkedIn, they can reach out to the entire people, you know, the Locus AI team. And at the same point of time, we also have a LinkedIn dedicated, you know, social platforms for Locus AI on LinkedIn and Instagram. And yeah, that's pretty much we do have a website as well. >> Awesome. I'm going to link in the description of the show notes. Thank you for joining the Locus AI team. I really the first step of growing your career in the space of AI is actually start building real life AI product with real users and ideally get some investors to reach out to you and want to give money put money in your hand and the standard one to get started is actually start building a real life AI product. So make sure to download the top 20 AI product ideas to gain real life experience. For people who want to uh join AI management boot camp which we assemble a team of soft engineer, developer and data scientists just like what we see today. Feel free to go to our website pmacerator.ioipm to learn more and feel free to also book a call with our team of uh product advisors to give you a free career evaluation to see if they qualify for the program. We're going to link it in the description of the show notes regarding how to book a call. Dear listener, if you find today's podcast insightful and useful and inspiration. So, make sure to like and follow and subscribe to our channel through all your favorite social platform such as YouTube, Spotify, Apple Podcast, LinkedIn, Instagram, Twitter and X and Tik Tok. So that you're going to get all the different inspiration help you grow continuously. And I also would love to have everyone to leave us a fivestar review. And actually I personally read all the reviews by myself and this is actually my big inspiration for us continue to create organic free content for you guys to grow the career in the space of AI. This is Dr. Nancy Lee from pmacerator.io and we're going to see you in our next episode right here and thank you for joining me Locus AI team. Thanks a lot for having us over