Building an AI Platform Company for Real-Time Applications: Vantiq CEO Marty Sprinzen (Part 5)

Sramana Mitra: “Operating system” isn’t the right term. Here’s my two cents—during the cloud era, platform-as-a-service (PaaS) became a well-understood concept.

In 2005, Salesforce launched force.com, enabling scalable applications and significant companies like Veeva Systems, which only had $4M in funding and was very cost-effective. They built a multi-billion-dollar public company on top of force.com. Later, they replaced the Salesforce stack with their own.

The idea of building on a PaaS is still not fully understood in today’s tech industry. So, don’t call it an operating system—it confuses people. Call it platform-as-a-service or PaaS.

Marty Sprizen: We do call it that—just used “operating system” with you since you have a technical background. Our platform runs distributed applications. We have our own event brokers and architecture for multi-tenancy. It’s similar in ways, but yes, the term can confuse people.

Platform as a Service is a little restrictive. I understand what you’re saying about Salesforce, but people don’t realize the extent of what a platform like ours can do. So, I’m open to any advice here on how to get out that message.

Sramana Mitra: You bring up an interesting point. PaaS came from the SaaS/subscription model era. But that’s shifting. Now, outcome-based pricing is emerging. Companies like yours, alongside system integrators and partners, are well-positioned to pursue this model.

Machinify, for example, did large, solution-level pricing—millions in revenue from single engagements. This isn’t subscription or services pricing—it’s outcome-based. Other models like usage-based exist, but outcome-based offers the most value.

Marty Sprizen: We do some of that. Another strategy we use is revenue sharing with customers. As they build applications, we take a small cut of the revenue.

We have one company with one of their products alone at hundreds of millions in sales.
They already have the primary product. So, we’ll get a percent of the future product sales. So that is not exactly outcome, as you said, although we have dealt with a couple of companies on that arena as well.

Sramana Mitra: That’s still outcome-based pricing. If you’re powering a product that generates revenue and get a cut of it, that counts.

Marty Sprizen: That’s our primary model.

Sramana Mitra: That’s great. So your innovation is not just in tech but in your business model too?

Marty Sprizen: Yes, we’re innovating on both fronts. Right now, we have no direct competition. Think of us like Palantir—but for real-time. Most players in this space are limited. We can integrate any number of sensors anywhere, with high reliability, scalability, and support for multiple LLMs.

We even have a patented voting algorithm to improve trustworthiness—if three differently trained models agree, it’s a highly reliable signal. We orchestrate those responses.

This is a new concept, and we offer a soup-to-nuts solution for building complex real-time systems. Demand is huge, especially over the last six months, likely due to generative AI. People are starting to ask: “How do we use this to manage real-world operations—things that we’ve never thought of before?”

Sramana Mitra: Are those use cases mostly in government—disaster management, traffic control, etc.?

Marty Sprizen: Yes, those are key early applications because ROI is strong. But we’re also being pulled into manufacturing, for example, because you can use a much more advanced system to manage everything from supply chains to inventory, to distribution.

You could really control these systems in real-time much more effectively. That’s where one of our large system integrators is building for a very large company.

This segment is part 5 in the series : Building an AI Platform Company for Real-Time Applications: Vantiq CEO Marty Sprinzen
1 2 3 4 5