Adept AI's Expansion: Securing $350M to Enhance AI Understanding of Software and APIs

0:00In case you missed the news, earlier this year, a startup called Adept raised$350 million. And in total, Adept has actually raised over$415 million. So today on the podcast, I'm going to be diving into what they're planning on using this money for, who their investors are, and what it is exactly that Adept does that makes it worth a staggering at least$1 billion. According to Forbes, they've reported the valuation of at least$1 billion. So this latest round of$350 million was a Series B. It was led by General Catalyst, Spark Capital. They had participation from Edition, Greylock, Atlassian Ventures, and Microsoft and NVIDIA, Workday Ventures, Katerina Fake, Frontiers Capital, PSP Growth, SV Angel, and A Capital.

0:51So they have a lot of big names in there. You're going to notice there's also a lot of very big software players like Microsoft, NVIDIA, and Workday Ventures, Atlassian Ventures as well. And then they have, you know, some of the biggest venture capital firms. So this is obviously a very high profile raise, a very high profile company. And so yeah, this this final kind of cash injection brings their total raise up to$415 million. And what's interesting is when asked about, you know, what this was going to be used for the CEO who is David Luan said that it's going to be put towards productization, model training and headcount.

1:31He said giant foundation models for language and for images have shown astonishing capabilities in the last few years. Adept is building on this momentum via a new kind of foundational model that can perform actions on any software tool using natural language. So I think obviously the word, you know, foundational model is a little bit jargony, but essentially what their vision is at a high level is that they're planning to create what is referred to as what well what they're calling an ai teammate so essentially what it's doing is this is not necessarily just a model trained to generate text or something this is something that is trained to use a really wide variety of different software tools and apis so instead of kind of going into different ways to generate text or images like open ai and they're more so just studying how people actually use computers and very specifically how they browse the web and how they navigate software.

2:22So the idea behind this is to help them train an AI model that they can use, you know, text instructions and turn that into sets of digital actions. So this, a lot of this idea is, came out of a paper that was published back in February, 2022 by scientists over at DeepMind, Google's kind of, you know, AI company, or AI backed, you know, company. So essentially, the paper covers the fact that they had an AI watch how keyboards and, you know, mouse commands were being used by people that were given instructions. So, you know, they would tell a person to do something like go book a flight or go learn how to do X, Y, and Z.

3:02And they would follow what like their actual clicks were and what they're like computer was doing. And essentially, they're training an AI algorithm, or they're talking about the fact that you could train an AI algorithm to replicate similar things. So the AI might not even know what it's necessarily doing. But you'd say, you know, when someone wants to do x, y, and z, these are the computer like clicks and the things they do. So it's a very interesting way of training and AI. And we also see at Deep Mind, the co founder Mustafa, he teamed up with the LinkedIn co founder, Reid Hoffman to launch inflection AI, which essentially aims to use AI to help humans work more efficiently with computers.

3:40And some people are saying in very similar ways. So just the fact that there is competition isn't necessarily scaring away investors. Obviously, we've seen, you know, they've raised quite a lot of money. And I think that this is probably due to the fact that there's a lot of market opportunity. So there was actually a recent survey done back in March, I believe. But essentially, they surveyed a bunch of AI professionals by Intel owned cnvrg.io. And around, I think, almost 50 % of them said that they believed organizations investing in AI development is going to increase despite, you know, kind of the macroeconomic downturn climate that we are seeing at the time.

4:21So I think Adept at the time of this raise, they were running a really lean team with just 25 employees, which is absolutely insane to have a billion dollar valuation with just 25 employees. But I believe since they've done this raise, they have increased their headcount significantly. So it'll be interesting to see where they're at right now, because this was a few months ago. So it's also reportedly that they are experiencing really high levels of turnover, which is not a very good thing for their company. And even that didn't scare away investors. But apparently they had two of their co-founders, Ashish Vashwani and Nikki Parma.

4:58They actually left and went and joined another startup, which, I mean, I don't personally feel like that is a very good sign for Adept. But once you've raised, you know,$400 million, especially when you only have 25 employees, I'm assuming that they're going to be able to have the capital to make something happen, whether they have to pivot or not. They have the capital in the AI market. They have some smart people. And so I think that they'll be just fine. But apparently, the two co founders leaving didn't actually disrupt the product development too much, which I think is good, especially considering the, you know, the high profile investors they have.

5:35So Adept's MVP is called Act One. And it can perform tasks like importing LinkedIn URLs into recruiting software. And according to Forbes, Act One essentially displays an overlay window on top of existing software like Google Chrome or Salesforce. So the prototype is ready for desktop, but will come to mobile in the near future. And essentially, just really the versatility of Act One evidently attracted some of these really strategic investors like Microsoft, NVIDIA, Atlassian, and Workday, because all of these big software companies obviously can see how this might someday benefit their own software and being able to integrate that into an AI assistant like this.

6:18And so I I think that's kind of what sparked a lot of these really big companies to invest into it, is they saw how this would help them. So General Catalyst's Deep Neshire said that Adept possesses a deep expertise to develop a commercial product that pushes the generative AI frontier beyond text and image modularities into the practical realm of knowledge worker actions. Excitingly, Act 1 has the potential to lower the barrier to entry within the enterprise workforce and thus may yield greater inclusive prosperity. So I think this is a really cool project. Obviously, they've raised a crazy amount of money and they're not without their difficulties as all companies are, but I'm really excited to continue following Adept.

7:03I felt like this is definitely one you need to know about as it is one of the top five, I believe, AI companies that have raised the most amount of money looking at strictly the AI and not necessarily someone like NVIDIA or someone that is benefiting from this industry, but an actual AI company developing their own AI technology in a really big way. So this is going to be a very interesting company to continue following into the future.