tektonic-ai

A platform developing GenAI agents to automate complex business processes, enhancing efficiency and decision-making.

Tektonic AI

Welcome back to AI chat. I'm Jaden Schaefer. Today on the podcast, we're talking about Tetonic that has just raised $10 million and they're doing drum rle pleas generative AI agents and specifically for the Enterprise. This is a really interesting area that is heating up and in my opinion is showing some of the most exciting innovation. I've seen some really incredible demos over on X of different agents people are building in hackathons. What they're able to accomplish and it really is making me quite bullish on where we actually go. 

This is part of the reason why I'm developing AI Box, which is a no code builder Marketplace because I believe this space is really really going to have a ton of disruption. So let's get into the [Music] podcast. The big thing that I think is happening right now with Tonic is that this is a Seattle-based company. So they are essentially combining general generative AI with a bunch of traditional kind of methods and they're coming out of stealth right now. This isn't something that has

Been around or launched for a very long time. They just did this $10 million was actually part of their seed funding round and it was led by 72 Ventures and Madona Ventur. So some solid people backing them right now. Their co-founder is Nick Santu. He previously had a bunch of different leadership roles over at Titanium, UiPath, and Microsoft. This is his latest exciting adventure that he's embarking on. So he really has been highlighted in kind of that the whole field right now.

In general, VI is unlocking a ton of really exciting different software scenarios that previously were completely impossible. Um, this is something I definitely agree with as well. I think this is really exciting when we start talking about automating systems and users automating what they do. He emphasized really kind of did his like press release thing and he was kind of talking about the importance of combining generative AI with symbolic methods and traditional software to really maximize your potential. He said quote.

You have to combine it with symbolic methods. You have to combine it with more traditional software if you want to squeeze the best out of it. This is really interesting because I've seen people have different approaches to gener Ai and even to agents. And in just like software and Ai and Automation in general, there's kind of like the one camp and really what we've done more traditionally, which is like let's say you're automating a car factory. You have these like robotic arms that do different things. 

There's kind of this new field and this new wave coming out. I'll share the robotics example, but it gets into software, so it's kind of the same analogy. But you know, we started out with these robotic arms moving stuff around. I feel like we're going to get pushed to a place where essentially we have humanoid robots that will do most of the tasks. Because when you're automating things that humans did, these humanoid robots are going to be able to do a lot of things better than automation to the point.

Where sometimes when you have like Tesla they over automated their car factory at one point with these automatic arms and they had to pull a bunch of them out because they just weren't versatile enough. And they brought in actual humans to do the job. Now those humans in my opinion are going to get replaced by humanoid robots. But I just think it's super interesting because we've kind of seen the shift where like we really have built the world from the perspective of like what a human can actually reach and do and move. 

And I think we're going to see, and just bear with me on this analogy, but something kind of similar with software where essentially we have these backends and these APIs and this code that kind of does tasks for us and we're trying to use it to automate things. But I think what people really want and what's going to unlock a lot of potential is for the people that actually do a task for that task to be automated in the same way that they do it. So you don't need necessarily a software.

Developer but you know an AI model that can watch how you use a computer. How you use software where you're clicking and so it's like simulating an actual human using a device. I think it's going to be a huge unlock and a huge way that a lot of our systems get automated. 

Now I know all the developers I know are probably screaming at me right now because they're like this is so dumb. Why are you automating it the way a human clicks on a screen and does things? You just get an API and there's code in the back end and it's like doing all this stuff. 

And while that might be way more efficient perhaps, I think if you do it the other way where it's like you're literally training an AI model to like watch what you do on a computer and copy that same like process flow keystrokes for example, I think we have a way bigger opportunity to train. Because we have a way bigger data set of humans that are using computers all day every day. And if all of them decide to allow an AI model to train off of what they're doing it, it's

Like we're instantly able to use the data and then we don't have to hire a trillion developers to try to always constantly change things. So I know these are like two like philosophically different fields, but in my opinion this is where things are going. 

And so I think it's interesting. It seems like Tonic is kind of going in that same direction. At least their CEO seems to be thinking that this might be the camp that he's going in. 

So right now I think while generative AI can definitely improve automation a ton, I think that there is definitely always going to remain a reason for a human in the loop for the seeable future. 

And I think that Tonic, they kind of believe the same thing. Their CEO said, quote, "Our approach will not make it 100% flexible for short. I'm not claiming that, but I'm introducing enough flexibility to cover a much broader set of scenarios than was possible before." 

So it's not like they're saying they're going to completely automate everything, but they're definitely taking steps.

In the right direction so right now Tetonic uses kind of a combination of a bunch of foundational models and open models for entity extraction and lower level actions. They're trying to become essentially a SAS company within 3 to 5 years that really is connecting APIs and various systems. Currently, the system is installed as a container in a business's virtual private cloud to get started. Um, and then it runs from there. 

This is really interesting. They're like they just raised this $10 million and they're really hoping that within three to five years they're going to be like a very viable product. This makes a ton of sense. Uh, these platforms people are surprised by how long these platforms take. I for one have been building AI Box for a year and a half. 

We're going to be launching our closed alpha very soon. You know we're shooting for the next month, but you know how development always is a little crazy. But it has been, you know, it these things always just take longer than you think and especially when.

You're making something really big really disruptive really powerful. So I'm really excited to see what they do with agents. It's obviously an area that I'm also focused on kind of building in the space with AI tools automation agents and kind of that area is what I'm looking at. I haven't, you know, we're not developing agents yet but it's definitely something that is on our horizons that we're looking at. This is an amazing area and I'll definitely keep you up to date on what Tetonic does in the future.

Note, that there may be slight differences in text and video. Watch video for the original content.

Tektonic AI Video