Venture with Grace
00:19:24 · Austin Spiegel
And I think these two things together certainly were a big realization within the business that AI is going to solve the business layer, or sorry, the application layer. The data model still really matters, right? So having, for example, in a CRM, having a good data model is super crucial because that's how you basically model all of your relations with all of your customers so on and so forth. But yeah, the visualization or the view, like we use Salesforce now, I haven't logged into Salesforce ever. Think I just use Claude. But in any case, so this kind of made us realize like, okay, the application layer is solved or going to be solved very quickly. So where do we take our business? And I think the thing that I still feel very strongly is that data infrastructure, so infrastructure that ingests hundreds of terabytes of telemetry per month, does that with sub second latency, is compliant FedRAMP, on prem, and has very strict access controls. These things are not something that you can just easily build with AI. And even if you can build it with AI, which we of course use AI to build it now, you still need a team who knows what they're doing, right? So I think that's defensible part of the business. And then we've been actually cooking up some really cool things on the visualization front where the principles apply, is, hey, there's actually a lot of complexity in visualizing high rate data in the browser, specifically like how the visualizations look and what the user experience of the visualizations is maybe more custom to each user. So how do we build infrastructure that allows us to visualize really high rates data in really performing ways while also giving the customer flexibility to visualize in the way they want to. So I think that's kind of like how do you make AI a tailwind for your business and then of course integrating AI into the product so that we can use agents for root cause analysis and so on and so forth. But then internally it's just like, adopting the tools, building, I think very popular concept right now is this concept of a software factory and adopting the tools to enable our engineers to leverage AI to move faster.
00:21:38 · Grace Gong
[00:21:38] S0: So one of the things that you said on a different interview or something is like, so hardware teams generates terabytes of high frequency telemetry per test, you promise like storing data stays queryable. So for life of the program without breaking a budget. So I guess how do you sure the economics works in this industry?
00:22:08 · Austin Spiegel
[00:22:08] S1: Yeah, good question. I mean, this to me is actually fairly simple. I think if you look at last generation database technologies, we were thinking about this all from the cloud perspective, right, because the cloud is kind of the easiest way to scale up. Now, I think actually that might all change in the next year or so because there's so much vulnerabilities being discovered by the kind of frontier LLMs that I think customers are going to be more want to do more on prem, which we already provide as well. But thinking about it from the perspective the cloud, if you look at all of the last generation of time series databases or OLAP databases, you have to store all of your data on disk essentially or EBS volumes. That's quite expensive and you're not really retrieving all of that data most of the time, right? Actually, think we found that upwards of 90% of queries are for fresh data in our platform. So that tells you that most of the historical data is ever being looked at, but our customers don't want to delete it because it's too valuable to them. So, then you need to figure out, okay, well, what's the cheapest way to store historical data? And maybe you're incurring a little extra latency, but that's acceptable because it's cheaper, right? And that's what object storage, whether that's S3, Azure, all of the major cloud providers leverage this. That's why we store all the telemetry in object storage. And then even the cloud providers also have intelligent tiering where you can actually move files to lower tiers of cloud storage that cost less but have higher latency. So that's kind of effectively how it works. Then the same just kind of applies to on prem, which is like, you could store it, you can either use some type of object storage provider on prem, you can use a distributed file store, kind of anything just to store those files. But that's essentially how we make it work.
00:24:06 · Grace Gong
[00:24:06] S0: Speaking of the economics, maybe we can talk about the life cycles of the company. Basically, your customer is in commercial space, defense and autonomy. Can you walk us through about the life cycle? So when does people typically plug in Sift and from early testing to manufacturing to live mission operations? When is it comes in? And in terms of the pricing perspective, how do you guys typically price the product as well?
00:24:43 · Austin Spiegel
[00:24:43] S1: Yeah, good question. So I think certainly earlier in the history of the business, we kind of came in post having some hardware stood up or some software stood up, right? Like, it's very difficult. You know, a lot of these businesses, they maybe spend their first year just, you know, designing, right? So clean sheet, let's design, let's define requirements. Worked with customers, we have active customers now who we engaged with a couple of years ago. They were creating megabytes of telemetry, and now they're creating hundreds of terabytes of telemetry per month. So, that kind of shows you how quickly businesses can scale, but also that there is a time where maybe it is a little too early. But now that we've built our edge application, which is not cloud, so we have like a full up telemetry database and visualization tool that can run on a Raspberry Pi, for example, collect telemetry with really high rate telemetry with minimal latency, I think that's actually really great for super early customers where they maybe don't need the power of storing all this telemetry in the cloud, but they just want the ability to run a test, visualize the telemetry, look for deviations and then potentially store it for the long term. So that has actually brought us earlier in the cycle. I mean, to answer your question, think better to get started sooner rather than later, right? Might as well use the best in class tools as long we can meet them where they are too, right? I acknowledge, I think I'm a big fan of startups. I love working with startups. They push us to be better. Also, understand that you can't sign up for an enterprise SaaS agreement or an enterprise agreement when you've got five employees and you've raised a couple million dollars. And I think it's really important for us be not only it's not only about making money, right? It's about being stewards of the ecosystem and helping us realize the kind of science fiction future that we thought we were gonna have. So it's important for us to meet the customers with a price point that's acceptable at those early stages and help them grow.
00:26:57 · Grace Gong
[00:26:57] S0: So I wonder, maybe we could talk about the building of the company part. Was like, you were the CTO before becoming the CEO lately. And I fact tracked, you did hire someone from Palantir as well. So it's like, you have hired exactly from and to worry, like Palantir, Applied Intuition, Meta and like, you know, SpaceX and like among others. So maybe we could talk about like, you know, when you were thinking about ensembling your own team, like, what's the difference between like at the beginning stage of, like, you guys were just building to nowadays, like, scaling, like, you know, for... You know, when you're hiring the executives from these large companies, like, what are... What is the core theme that's or core business milestone that you're trying to achieve for, let's say, between now to CRC?