I saw an article recently about the “forward deployed engineer”, or FDE. AI companies are building enterprise units that embed directly with customers to help them figure out how to make AI actually work inside their organizations. The reasoning is straightforward. Customers aren't sure what they need from AI. They're not sure how to deploy it or how to make it successful. So the software company sends in a team to work alongside them.
It's a good model. But it's not new.
Palantir did an early version of this 16 to 18 years ago. They deployed what they called echo and delta teams. Domain experts paired with engineers who were comfortable working without firm requirements, figuring out solutions quickly, directly alongside the customer. The goal was to understand real customer needs and deliver working solutions fast.
When I read about this, it reminded me of how Kodaris started.
Back in 2013, when Kodaris was a one or two person operation, I was personally embedded on site with our first customer. I worked with their different departments. I learned how their business operated. I went with them to visit their customers and learned how those customers did business and how they interacted with our customers. All while we were developing the product at the same time.
That went on for 4 years, off and on. The result was a product with strong market fit because it was built in direct contact with the people using it.
Kodaris does a version of this today, mostly remote and across a much larger customer base. We have weekly meetings with customers with real conversations about what they're adopting, where they're struggling, what problems need to be solved next. We release updates every 7 days and work from prioritized roadmaps built around real customer problems.
It's not the traditional SaaS model. Traditional SaaS is built around self-service. You build the product, document it, and let customers figure it out. Support is reactive. Roadmaps are driven by internal planning. Engineers are separated from customers by layers of product management and support processes.
Kodaris is a hybrid. We have a self-service platform. Customers can use it without hand-holding. But we also have engineers working closely with customers on new products, enhancements, and brand new modules every single week. As the Kodaris community grows, that scales too.
The pattern that keeps emerging in technology is that the methods themselves don't change as much as the tools do. AI has been around for decades. What changed is the capability and accessibility, not the underlying principle that understanding customer needs produces better products.
The same is true here. Embedding with customers to understand real needs before and while building is not a new concept. It's just not what most SaaS companies do. The economics of self-service scale are too attractive. Build once, sell many times, minimize services overhead.
But the companies doing the most interesting product work are the ones staying close to their customers. The AI companies doing it now are doing it with large enterprises trying to figure out how to operationalize the technology.
Kodaris does it with distributors and manufacturers. The niche is different, the scale is different, but the model is the same.
One thing this helped clarify for us is a question Kodaris has wrestled with for a while. Are we a SaaS company? Are we a services company? We do a meaningful amount of both and it's sometimes difficult to explain.
The forward deployed engineer model gives a useful frame. Kodaris is a platform company that stays embedded with its customers, because it's how you build the right product. Working directly with customers, understanding how the core platform needs to evolve, and releasing that evolution every 7 days is how the Kodaris SaaS grows with every new customer and every new subscription.
It will be interesting to watch whether the FDE model becomes mainstream over the next 3 to 5 years. My guess is it will.