E3: Reviewing startup pitches from cancer cure Granza Bio to Brazilian Telco Salvy | YC W24

Welcome to the Superconnector podcast. I'm Matt Joseph your host. Today is a special show because YC demo day is April 3rd and April 4th, 2024. The winter 24 batch will be presenting. And I'm excited to dive into some of these startups that we'll be presenting. We'll hit probably like the top 10, maybe a couple others that we're excited about.

If you don't know what demo day is, I once described this as a beauty pageant for startups, I think that still applies. Essentially all of the startups in YC will present except for some which have already raised their money and decide not to, but that's a different story. They will present to investors, investors who are excited, can reach out to them.

There's also an alumni demo day. So an alumni demo day, people who have already gone through the program get a chance to reach out to the startups [00:01:00] preempt rounds or join rounds that are already being raised by them.

And in honor of YC demo day, I'm going to run through the top most upvoted startups on the YC launch website. That is ycombinator.com/launches and you can filter from there.

So let's go.

Startup number one is Granza Bio. Cancer targeting attack particles derived from our body's immune systems.

All right. So let's click through on this and see what we got. To PhD graduates and oncology and immunology from the university of Oxford on a mission to develop personalized and targeted cancer treatments.

We have discovered a new weapon in our immune systems arsenal called attack particles.

We are turning this concept into a therapeutic product that can eliminate any cancer. Without harming the surrounding healthy cells, preventing adverse side effects. [00:02:00] And I guess they're both of their names are Ashwin. That's kind of cool. We've got a nice photo of them hanging out outside of the YC office. Little bio on them, but let's jump to this video that they've got here.

cell phone. I've worked in some teams, but not all. These treatments. The surrounding healthy tissue causing high toxicity. And adverse side defense.

We are developing attack particles that are engineered to specifically target cancer cells. These particles are inspired from our body's own immune system and are composed of a thrombosis bonding protein shell loaded with cytotoxic molecules, like granzymes and perforins. Okay, I'm going to pause there.

That sounds really cool. I like it.

It's way above my head. I mean, this is like one of those bets that you do because you just think it's really interesting. And the founders are super smart. And you hope that they figure it out over time. This is not [00:03:00] a business that can get built very quickly. But I think it's cool.

So startup number two, Salvy, simplifying teleco for businesses in Brazil. Let's jump into this. See what we got.

We allow more than 250 companies to easily purchase and manage their mobile lines while reducing costs. Okay. So an international team. Focused on the Brazil market. I've heard great things about Brazil.

Customer pains, according to them, cost inefficiency. No transparency. Companies have no autonomy to manage their lines, awful customer service, confusing payments. These are all things which you hear often about the LATAM markets. And it's a hot market right now.

Let me talk about this business a little bit. So this is part of a bigger trend. And if you guys heard the episode I did with Joel Karacozoff , who's invested in over 40 YC startups.

It fits right in line with his thesis which is India and LATAM are exploding.

I think geographically our group has conviction that there are [00:04:00] areas around the world where tech innovation is going to move faster than in other places. And India and Lat Am has been one of those, and again, very agnostic to what it is they're building. It's just the opportunity there seems very, very ripe.

Really all the emerging markets are great to invest in right now.

I think this team, it's hard to know how well they're going to be able to do, but I like what I see out of them. They obviously attracted a lot of attention on YC.

And of course there is an AI component to this. We use AI to read telecom invoices and help companies migrate to Salvy. In just over a year we've acquired 250 companies and have a 322% NDR with 0.2% churn. Well, it's cool that they have figured out how to use AI to take other companies, customers. Maybe that's one of the killer use cases for AI.

I mean, we're going to talk more about AI a little bit later, but. Cheers to Salvy. Congratulations on launching.

Okay. Number [00:05:00] three, Hona. Less time with charts. More connected patients. Okay.

This one, I actually know a little bit about, I don't necessarily know about the founders or the startups specifically, but I know about the space. So let's dive into this. Recent regulations have enabled electronic health record systems to communicate more effectively. However, healthcare providers are not adequately equipped to manage the substantial influx of unstructured data resulting from this enhanced interoperability. The average patient record is now the equivalent of half the size of Shakespeare's Hamlet. And providers often see more than 20 patients a day. For each patient, these providers find themselves contacting other organizations to obtain the patient's history.

Then they must sift through numerous pages of PDFs to locate critical, relevant information. Similar to finding a needle in a haystack. Wow. What a reference. This process creates an enormous cognitive [00:06:00] burden leading to burnout, missed revenue opportunities and most critically uninformed care. Okay.

So this seems like it was written by non-doctors. And let me just check and see if any of these are medical doctors. Doesn't look like it. Hona was founded by ex biomedical engineers seeking to return to their roots. Drawing from our experience in early stage startups and launching AI products for millions of users in big tech, we ventured back into healthcare. All right.

So we have three co-founders who are, I guess, big tech co-founders who are trying to come into healthcare.

Let me say what I really like about it. And let me say what I think the challenges are with doing a business like this. What I like is that I think they are onto a very real problem. These charts have gotten out of control, the electronic medical record companies like Epic, Athenahealth. There's a whole host of these guys who are floating around and the charts are enormous.

They [00:07:00] really are. The files are giant. You can imagine. A combination of x-rays and charts, notes that the doctors are taking along the way. If the patient has been to multiple facilities all of that is getting consolidated into this chart and the chart can be very complex.

So I think these guys are onto a pretty important problem. The challenge I see with a business like this is that you have to get the electronic medical record systems too agree to use this. And I think epic owns something like 53% of all hospitals around the country.

These are like effectively monopolies in the facilities that they operate in and getting the hospitals to make changes to them is an absolute nightmare. It's extremely difficult. Like this is a business which is going to be dependent much more on relationships and ability to sell than it is on the intrigue of the technology. I would expect them to build [00:08:00] integrations with EMR, but then the only way to get your integration to be live and to be approved is to then have a bunch of people in the hospital sign off on it and that takes a ton of time. So I really hope that they're good at sales. This is going to be a very sales, heavy business. The other challenge you're going to have is that when something goes wrong, let me paint an example here. Patient comes in. Patient is having, let's just say a knee implant put in.

So they're dealing with an orthopedic surgeon. The orthopedic surgeon looks through the chart. They use Hona. And they miss something critical. Patient has some sort of issue. Where is Hona in this now? How do you deal with the assignment of risk and liability for things going wrong? going to be a huge problem for this business. Maybe they've already figured it out.

It's hard to say. I'm assuming they're going after small providers to [00:09:00] start and they're going to work their way up to large guys. I hope that they figure it out. I think this is a cool business. Just going to be difficult.

Up solve. The new way to build an offer analytics to your customers.

All right, let's jump into this one.

Build customer facing analytics that get your users to take in product actions. And become obsessed with your product.

All right. So they have a little video here. Let's see what they say. Hi, everyone it's Ka Ling. Today. I'm going to talk about how companies use Upsolve AI to build and offer analytics. Okay. So for our listeners, she's got a really sweet hat on. I mean, you can't see it, but. It's like to that customers. It's like kind of striped.

I mean, you know, on that alone, she gets points for me. Let's say I'm a product manager of this marketing automation company. Everything using ups. To build a sales performance report for our customer. But [00:10:00] recently we got a lot of inquiries. Well, her hat actually matches the graph that she is showing on the screen. Cool lighting. And just total sales KPI. Smart. They would love to see in a geographic adventure. Doing it with ups off is extremely useful. We can explore further. Six sample. I don't feel this. Okay.

So let's like find a use case here. For example, marketing companies use up solve to help users optimize their campaigns and FinTech. Companies use it to help users manage their accounts. All right. It's a little vague. I'll be honest. Like this doesn't seem like it's a

particularly fleshed out use case yet. I mean. Speaking from experience selling to marketing companies and selling a FinTech companies is a very different sales motion. So they're kind of just throwing shit at the wall and hoping it sticks right now. But, like I said, she's got a sweet hat and I am hoping that they can succeed.

At [00:11:00] the end, they say, if you know any B2B business in your network, that's looking to unlock customer facing analytics and reporting capability,

we would greatly appreciate any warm introductions. A lot of YC companies have done this and so. I'm intrigued to see like what their go to market is like how they plan to attack the space. But. Good luck. And sweet hat Ka Ling.

Fume. AI SWE, that solves bugs within slack. Okay. What do we have here? A slack bot. You can assign software tickets and get full PRS in response.

All right. So. Basically, this is part of this trend where software engineers are being turned into AI, or maybe. Startups like fume are basically trying to take pieces of what software engineers do. Co-pilots, there's been a lot of those YC has [00:12:00] funded a ton of them.

Doesn't really seem like there's necessarily a secret sauce here.

This is execution driven. They want their software engineer to fix bugs, error logs, et cetera. Non-technical team members can interact with it and it's integrated into slack. So, you know, the idea is that it would be pretty simple to start using it. It sounds cool. Let's move on. Good luck fume.

Math Dash. Turning math into a sport now. Now we're onto something interesting.

Not that the others weren't interesting. But. Math as a sport. This makes me think of just e-sports as a whole and the explosion of that industry. I'm excited to see what these guys have to say.

Competition math the best way to learn math. All right. Competition math involves solving challenging and creative problems within a competitive setting. Covering topics beyond standard school curriculum. This format sends competitors to the frontiers of math and science by [00:13:00] encouraging innovative thinking, fostering a community and promoting a love. For math. That's cool.

So basically they're making math competitive. They have this whole idea of mathletes that’s kind of cool. I don't know what’s unique about this as compared to some of the other things that are out there but I guess it’s just that you can play anytime you want like you can just jump into any competitive math environment. OK so here’s traction. We have 200 of the best competitive mathletes participating in our tournaments and our users are solved over 400,000 arithmetic questions. One even played in our arena for 10 hours straight. Well you got a lot of math lovers here That's cool.

Okay. Moving on. Paradigm. Your AI intern to supercharge email outbound. I hope you guys are noticing the number of AI startups here is very, very large. So [00:14:00] let's see what paradigm's doing. It's funny, like AI intern. What they're doing there is minimizing your expectations, right? So they're like they're calling it an intern because you don't expect your interns to do very much, which will allow them to eventually turn it into a full-on sales development rep that is operating at crazy efficiency, but they are tamping down expectations which paradigm I like good job. Automatically reach out first with spot on messaging that guarantees a response and a sale. Well, now that is a bold claim. Miss Monaco, Anna Monaco and Claire from paradigm. We are obsessed with building AI agents that complete your boring tasks.

We think of these agents as able to handle the most repetitive parts of work while iterating with your feedback, just like an intern would.

I'm assuming that they're relatively young because I. I expect all of my employees to iterate with my feedback, not just the [00:15:00] interns. But point is well-taken. Excited to share a glimpse of our sales use cases. All right. So I hope these young women have actually done a lot of selling themselves. Anyway, I'm going to read out what they have TLDR. Paradigm 10x's your sales funnel by number one, getting hyper-specific leads for potential customers with active buying intent. Number two, writing them personalized emails that sound like they're sent by you, not a robot.

Number three, letting you approve and edit emails before sending them or putting it on autopilot. Number four, tracking what works by integrating with CRMs and spreadsheets and providing tailored analytics.

Okay.

I want to see sales fundamentally change. This to me does not fundamentally change sales. It just means that I'm going to get more shitty spammy emails in my inbox. I don't think that's going to make me feel any better. I already feel like those emails piss me off. Like it's hard to cold email somebody with relevant information. I think that requires a lot of [00:16:00] research. It's not just about personalizing the email, it's about pulling in information about that person that you would only know if you knew them. Right?

The sales email that I'm going to respond to is one that's like, Hey dude, you've been sending out lots of emails to podcast guests.

Like in YC, we know that you're doing that. And we have a better way to do that for you. That's an email that I will respond to.

But.

As with any startup it's early.

Yeah, I figured. It's exactly what I thought. We have been building together for the past four years after meeting in Penn's infamous, discrete math course.

We started our first business together two weeks after the meeting, a software dev and design agency, we bootstrapped nearly seven figures in revenue. All right. Like cool.

Software engineer and PM at Google and Microsoft. Yeah.

They are young [00:17:00] ambitious people. Good luck to the paradigm team. I would encourage Paradigm to rethink what sales ought to be all together. Like anyone who does a business like this is just trying to make money and I get it. It can work, but like, Do more.i As my advice.

Take it with a grain of salt.

Good luck paradigm.

Next up Apriora.

All right. I liked the name. Like a priori, a priori. I like that. You are AI interviewer to identify the best candidates faster. Reduce recruiting overhead interview more candidates and make better hiring decisions. All right. APR AI interviews, job candidates in real time companies using apriora. Blah, blah, blah.

We've already covered that. Interview processes are slow, expensive companies spend $5,000 per hire. [00:18:00] Most of which is time spent by internal recruiting teams. For processes 30 days. Okay.

Alright. Our AI. Interviewer conducts real-time interviews with candidates, including technical screens, phone screens. System design, coding, behavioral, and more. Unlike interviewing your candidates manually or with an automated assessment it's convenient. Reliable. Okay. So let me. I'm assuming that this is like a video agent.

Real-time video interviews with candidates.

Yeah. So this is like something I need to see an example of like you guys got to okay. Check out full demo at a priora dot AI slash demo. Okay, here we go.

Alright, let's get started.

Hey, Aaron. Thanks for joining me today.

It's nice to meet you.

I studied computer science at brown university. [00:19:00] I'm really passionate about software. And all things. Oh, my God. So really excited to. Oh, okay. Hold up. I need to fucking pause this. That's great to hear it. Okay.

Okay.

So for those who are listening and not watching. What's happening here is you, the candidate are speaking to an AI that you can't actually see. It's like a little box in the bottom right corner. You can see a live transcript of it. This feels to me like something that would not move the needle in the world of recruiting. Certainly I would not use it to hire people for my company.

But maybe there's people out there who feel differently about that. It seems to me like this use case is not for small companies. Like it's really only big companies with huge recruiting departments where this becomes an issue.

And for them, they have the resources, like cutting costs with something that hurts the candidate [00:20:00] experience because they're like, why am I talking to a robot? That's not great.

Not great. Have these guys ever interviewed people before I can't tell. I don't know what their backgrounds are. They don't list them.

Okay. Well, moving on. Good luck Apriora, I still love the name.

Next up Momentic. Automated testing with AI. All right. Another AI company.

Use AI to automate the most tedious parts of software development. Well, this sounds. The hell of a lot, like the last startup that was using AI to automate software development. Testing still sucks. Okay. So these guys are focused on testing.

That's kind of a good use case.

What if testing could help developers ship faster? Our low code test editor replaces traditional test scripts with AI powered self-healing steps. It seamlessly integrates into your existing workflows from local development to CII CD environments and production. That's continuous integration, continuous [00:21:00] deployment for those who are not familiar. And they have some really nice little graphs.

Yeah. I mean, okay. It's technical. These guys seem smart. I think again, this one comes down to execution. It's hard for me to tell exactly how strong the execution is on this. From what I can see. I think it's something you'd have to play around with, but again, Agents replacing software engineers is one of the big waves of the future. Momentic. Good luck. I wish you well.

Okay, next up HostAI, the operating system for vacation rentals. We provide an all-in-one platform that helps vacation rental managers. Deliver five star guests experiences at scale.

We are pulling and coal. The founders of host AI Poon. X school engineer and prop tech, builder, Cole, vacation, rental manager, and real estate developer. That seems like a [00:22:00] good pairing of a team. Like, you know, a guy who like knows how to write code and a guy who like actually managed real estate. I hate that AI is getting shoehorned into everything, but like that's where the cycle is right now.

Vacation rental management is hard.

Oh God. Okay.

Yeah. I have seen this concept attempted. So many times. And I think it comes down to how much are you willing to brute force the sales motion of this? There is a large graveyard of companies that have been like, we're going to solve vacation rentals. Yeah, like it's hard. Not just because the technology sucks.

It's hard because it's fucking manual. It's like a very manual job. People tend to try to just outsource it. They just want to hire somebody to take care of it. So, I don't know. I think it's tough, but. These guys seem like a good team. I hope they figure it out. [00:23:00] Good luck. Punn and Cole.

All right. So how many have we done now?

We've done 1, 2, 3, 4, 5, 6, 7, 8, 9, 10. We've done the top 10. All right. So I kind of want to do a bonus one here just to like find something cool to talk about.

I mean I kind of liked this. Garage.

Helping fire departments buy and sell equipment. Garage is a marketplace for used equipment, starting with firefighting equipment. I mean sweet. They got a picture of this kid. Little black kid in a firetruck. I'm sold. where do I invest?

Garage founder's hit me up.

Oh, is this a solo founder or, or no?

2. Alaz.. Met while studying in Columbia. All right. Firetrucks are expensive. Used trucks range from $50,000 to a million dollars. They're helping them buy and sell used [00:24:00] equipment in seconds. Payments, freight, financing. I mean, cool. I don't know too much about it, but this is a sweet picture. All right.

I think that's enough of YC for now. Demo Day is coming up.

Good luck to all of the YC startups.

‍