Conor Brennan-Burke, Hyperspell: Building data infrastructure and AI tools for the app layer

welcome to another founder wisdom B we have Connor Brandon Burke with us today he's coo and founder at hypercell rag as the service whatever that means we're gonna explore that meaning together so X A4 Capital so he did some Venture scouting in there and agel investor uh founder fellowship at odf 21 on on deck that is one of my favorite uh I describe it as an incubator

sort of I founded a bunch of companies included able and some others that we'll talk about today uh won School educated guy so pretty fun guy we had a pre-c conversation about SF and AI we're certainly gonna talk about that today pod py is presenting that podcast that is my podcasting company if you want to start scale monetize the podcast through YouTube AdSense lead generation and sponsors

can go to pod.com we have a free ebook there and a master class to accelerate your podcast learning Connor welcome to the Pod tell me more about yourself and what you're up to nowadays thanks charl it's great to be here um excited for this chat uh yeah I think you gave a a bit of a sense of my background I'm originally from Western New York Rochester went

to haford for undergrad while I uh did an exchange at Wharton um started off my career in uh Consulting working with big companies helping them transform their operations with tech and just got really really interested in building and software um after that I worked at checker for a while it's an API for background checks uh was able to build apis that and background checks products that were

used by Uber and door Dash and instacart and other big tech companies um started my first company uh AEL genen which was um focused on AI music so actually building AI music tools uh for producers and DJs we got that to to profitability and bootstrapped it and that led me to my current Venture which is hypers spell so hypers spell is a building infrastructure for the AI

app layer we allow you to connect any data source and make that data useful for AI right um okay so let's start with uh able Jen uh you did that for one year and you targeted music producers tough audience right so what did you learn from that one and uh yeah why did you close it down yeah so I actually didn't close it down um it's still running today uh shout out to the team that's uh working on it rebranded as Tona AI uh it's an incredible team and they're still operating um I found it challenging to run two startups at once and so I've moved into an advisor role for that team but uh it's continuing going uh yeah that yeah yeah Eble Jem was a um a great experience um so it actually

was born out of a hackthon uh and so we um you know showed up to this Ai and music hackthon and I didn't have a team uh and didn't have anyone to work with I kind of went on a whim because I said I've been a producer DJ been interested in music all my life I love AI um well I was at the hackathon I met some incredible folks F and we decided to um we were coming up with ideas and kind of jamming and sitting at the table together and then it said hey why don't we form a team and start working on this um and so you know it was it was quite an experience it's actually my first hackathon I've done since you know freshman year of college and uh the I

think everything that could go wrong did go wrong you know we were building a AI music product and uh we um the judges came over to us and uh said okay great we want to see it you know are you guys ready and the product wasn't ready yet so we said hey you know give us a second we need to restart the laptop and then like ran out of the room and you know it in the hallway and then came back in and they said all right like great we want to hear it now what does this AI music sound like and so we we got a prompt from the judges and then uh and then we enter the prompt in and we pressed play and no sound came out it was it was really

awkward right there was just silence then we pressed play again and you know no sound and after a bit I was like well just imagine that you were hearing music right now let's just picture how that would be you know um and what a cool experience that would be and then we realized actually that uh you know we were trying to fix the code and figure out

as a problem with the program and then we realized that the speaker we were using it ran out of battery and uh so you know was able to play the music eventually uh yeah and then uh we ended up winning that hackthon you know we were uh the top five teams at a a nightclub in SF and you know they were announcing the the third place team was this team second place team was this team and the announced us as the first place and I think it really for me you know is really validation of our idea and what we want to build and that there was market for this a bunch of folks reached out said hey we're interested in investing and our team said you know we could go the Venture back route

but we want to build something for creators and producers and Builders and uh let's just bootstrap it instead and build it from the ground up and build a a sustainable business that is artist first and focus we don't need external funding for that and how does that one compete with Sunol yeah so the it's a little bit of a different Market you know what we realized was

I've been a music producer for years and uh you know DJ played a bunch of instruments um and the tools that you use for creating music you know I was having this experience where in my day job I'd be working on B2B SAS and you know creating these beautiful uis and thinking about the customer experience and then in um and the I would be producing music and

I I felt like going back in time 30 years you know use this clun interface and uh really like uis that are hard to navigate it's not streamlined Bas things that we kind of take for granted like real-time collaboration or you know saving to the cloud or other you know kind of basic things in software don't exist in music production right now like if you're working in

uh Ableton Live on your computer and you want to collab with somebody else you're sending files back and forth and trying to collaborate there's nothing like Google docs for music and so uh and there's a bunch of other things that were kind of uh nent and not easy to use and so um for me I said well you know I'm uh there's an opportunity here toso is

great if you're consumer and you're just someone that you know likes making music and you don't have any musical talent but for all these DJs and producers what if we could embed AI into the tools that they're already using right you're already using this workflow that's kind of outdated these companies are never going to develop AI features on their own and incorporate it what if we were

that bridge where we can take all these Innovative new AI music tools and then bring that in and so yeah that was kind of the Innovation that uh that sparked able genen of how do we bring AI to people that are very serious musicians right and how did that lead to hyper spell what uh are you seeing in uh one integration and all the world's data yeah so um I I think they're very two very different companies um very different sides of myself right uh I've been again playing music for uh for years and it was something I was passionate about and interested in and I learned a lot in the experience of building able genen now Tona of you know we we were fully bootstrapped so as we were recruiting team members we said

well we can't pay anyone um we don't have any Capital so let's figure out a structure where we can just pay people in equity and profit sharing and then recruit a team there and today there are 10 people working on um Tona the new company with all page just equity and profit sharing um so it certainly taught me a lot about how to recruit how to hire

how to build a team how to motivate folks how to get customers but without any of the normal trappings of VCS and accelerators I think for me I just realized that there was a much much bigger opportunity in building something like hypers spell and uh you know I love music and I'm passionate about it but I wanted to swing for something much bigger um and so for

hypers spell we started off building an AI agent to automate your busy work when I was a product manager I spent a lot of my days kind of copying and pasting information from one format to another and I said well there's got to be a way to have an AI solve this much better than I did uh and so we started building that AI agent and then we realized for that to be useful we had to connect to all these data sources so we had to connect to your Gmail and pull in information there and we had to connect to your slack and pull in data from that we had to connect to your Google Drive and learn the context about your organization and really quickly we looked at all the tools available and we

found there's really nothing out there that takes data and makes it useful for AI and that you can easily connect to you know there's a lot of kind of Point Solutions and tools to take pieces of the pipeline but nobody had actually said hey let's go end to end and take care of all this for you uh that's how hypers spell was born so you do consider it as a moonshot um what could emerge out of this what would be like a final is evolution of it do you think you could could make a an AGI for work in there yeah well you know we're building tools for other builders so what we're really focused on is the next generation of AI Builders and um we're excited for somebody non-technical for example to say hey

I want to build an app where you can just chat with your email um and you can spin that up in two seconds right where instantly you can talk to all your emails you can ask questions you can have a auto reply to things um or let's say you want to uh you know build an app that um summarizes your slack data or it looks at your Google Drive and you can chat with it and interact with it you could build an agent that can interact with all these systems what we're really excited for is to empower this next generation of Builders and and what we believe is look the future of work is AI there's going to be more AIS working than there are humans and we want to build the layer to make

it so that all these AI can work with all these tools and data that already exists so sort of building the roads the high infrastructure like strip did for online transactions um why is zap you're not building that and how are you competing against replit and bolt yeah so um the way that we think about it is like uh there's are you familiar with a company called

plaid bad plaid pla yes yes yeah so what plaid did was um there was this emerging fintech ecosystem of uh okay all these companies you know Robin Hood venmo all these companies that wanted to build fintech but all of them needed to connect to bank data in order to do that and the bank data was kind of the source of Truth and all those companies could have

built their own connections you know Robin Hood could have built a custom connection to JP Morgan and all these Banks but all those companies said hey we're experts in you know if you're Robin Hood we're experts in stock trading if you're venmo we are experts in allowing you to pay each other we're not experts in the underlying infrastructure um and the exciting thing that pla did was

make it so that all these companies could build on top of them and focus on the things that they were good at and the things that were their domain expertise and that their customers cared about instead of this underlying infrastructure um we want to be the same thing for AI data um and so uh you asked about you know companies like repet other tools like that I

think we're in a bit of a different category there um we would consider ourselves to be much closer to you know raggi veara uh some of the point tools like unstructured index um with the key thing that's different from us is we'll actually go end to end and solve this all for you so you can set it up in a few minutes and get to high quality data instead of spending time using some of these other Frameworks and connecting them and who's your ICP do you target technical people or not not technical ones yeah right now we're focused on technical Builders um so we are focused on if you are someone that's building an AI and you are spending all your time trying to figure out how to do data chunking how to parse PDFs

how to um think about how often you refresh data that you're ingesting from slack how do you look at markdown in documents and then translate that into useful input if you're trying to figure out uh how often you are ingesting data how to store it if your system breaks every time you add a new customer and every time you try to add a new data source then

hypers spells for you um if you want a tool where you could just pick any data source specify hey this is what the output we're looking for is um and then focus on your enduser application that's really the problem that we're solving love it uh you got some funding so preed I would guess what are your top goals for this year at hypers spell our number one

goal right now and always is serving customers and so uh you know we want to make the reason we're doing this company is because we are really really excited about this AI wave that's happening and we want to empower other people to build you know the reason we started the company was because we kept talking to Founders in Ai and they all faed the same problem and

they were all trying to build this data stack and they were all trying to think about hey do we do this in-house do we use um external tools how do we string them together and you know for us we were just excited about all the things that all these people were building and we wanted to empower them and make them happen so you know we're we're going to be hiring uh this year with the funding that we've raised um but our real focus is just we are so excited exed about this new wave of AI companies and how can we help support as many of them as possible and is it paper usage how much should the technical person paid to use hypers spell yeah for our ear customers it's actually um we're giving them

unlimited queries uh so you can use it as much as you want which we think is a pretty incredible deal but we know a lot of these other AI companies are also very early and so you know our our thing that we offer to them is hey you pay one flat fee we will handle your entire data pipeline we will connect to the data sources we will figure out off the counts we'll figure out how to manage permissions we'll figure out your chunking strategy your parsing the embedding even the type of queries we run multiple queries at the same time all you need to do is tell us here's the data source and here's the type of output I would be looking for and then we guarantee a quality pipeline in the middle we charge

one flat fee per month um and then we give you unlimited queries based off of that right and you also do some Venture scouting how does that work is it like full commission and what percentage of your time does that take yeah I I started um I started doing some Venture scouting just because I was you know in this role at hypers spell I was meeting a

lot of great Founders and chatting with a lot of great Founders and meeting a lot of folks that were building incredible products that I was really excited about um and then on the other side you know we had been funded by some uh um some investors that we really really liked we thought added a ton of value and were extremely helpful and we were grateful for for

uh having them Believe in Us early on and so it was kind of like putting two and two together like all right well I know these incredible Founders um and I know these incredible investors let me just introduce them to each other and uh you know um a four Capital was the first one to uh invite us to be a scout where um they had uh they had funded our preed and uh invested in us and um we were able to connect them to some some of our friends who are also building uh incredible companies and then uh I I think it's it's very much something that was happening anyways is just putting a label on top of it you know uh it's very lucky to know both great Founders and great investors um and

then also recently joined the weekend fund as a scout uh so the weekend fund is Ryan Hoover founder of product hunt and that was very similar right um I met via uh on the team and Ryan and I I thought hey these guys seem great they they're um they're Builders themselves they've built product before they have a product background they offer real value uh to their Founders

and then I had some founder friends that were uh building things that I was really excited about and just started connecting them and uh pretty soon that turned into a scouting role at the weekend fund as well um so feel very lucky to have this role you know it's h everybody wins right the founders are excited the investors are excited I'm excited because I was able to

help connect them and so it's um it's been a lot of fun to be able to give back and give back to some of our customers you know we we have customers that um we've been able to help connect them to investors and uh and that's that's really nice to be able to serve them in that way right then you're also an angel investor what is your investment thesis on the startup founder side of things and on the startup itself yeah so I think I I invest at the precede stage um and at the prip it's really about the team it's it's really really about the team because teams pivot all the time and uh and you know we pivoted ourselves right we were building an AI agent then pivoted to building infrastructure my my

thesis is look for people that are doing very non-traditional things since young age um they have some reason uh something that they want to prove you know they have uh they're Outsiders that have kind of found their way in um they pushed through challenges in their life to get to this point you know uh one of one of the founders um I backed early was a 18-year-old

who dropped out of high school to start a company and um and going against consensus going against the Norms uh I also I one thing I optim for a lot as teams and you want a diversity of skill sets on the team you want an incredible technical team but you know I I asked them questions like what's the best argument you've had with your co-founder what does

that look like when you're um arguing with each other and you have different points of view how does that end up um what I optimize for is not how long the founding team is known each other it's can they have really difficult conversations and still be friends afterwards and there are people who have known each other for years and been friends but they've never had a tough

conversation and so they actually are never able to do that and um and there are people who uh have known each other for not that long of a time but they've created a great Dynamic and they can really push on each other so um on the founder side you know I I think it's like deep domain expertise a why for this some deep sense of motivation that um will make this person work on this problem for the next 10 years and then a really really cohesive founding team that I feel like can work together um I also like to see that Founders are thinking about the upside and the downside you know one of my favorite questions to ask is like okay tell me a story of how this is going to 100x the vision

that you have right now okay your vision is here how do we get even higher what does that look like and then tell me a story about if this doesn't work out why won't it work out right like what what's the biggest risk right now um and I find that gives you a lot of insight into the founders mindset on the startup side you know we invest a lot in the AI app layer just because that's who our customers are um the the there's a lot of plays there I mean very excited about like traditional Industries uh and so I think there's an opportunity for you know there's many industries that never adopted SAS to begin with because they just never got used to using a UI but they are used to talking to people

right and so if you can have an AI agent that they're talking with I think they can actually Leap Frog the previous generation of SAS um and then I'm also really excited about other infrastructure tools I think we're very much in the we're we're not at the iph era of AI yet I think we're in like the PM pilot era I think there's a lot to come in terms of what we're building and what's going to happen and so uh I'm excited to find folks that are saying hey we have this incredible um tool of abundant intelligence and how do we apply that to a un unique domain and make sure that it uh intersects with reality Connor I have so many other questions for you but we're already on time so let's do a

a round one next year to follow up on what you're you're up to till then uh where can people find out more about you yeah you can find out more about us at hypers spell.com that's h h y p r p l.com um you can find out more about me on uh LinkedIn and Twitter um happy to chat with anyone if you're building something you're really excited about an AI feel free to reach out that was Connor and Charles for another founder wisdom