Moderne is building automated code remediation for complex code bases

9 months ago 398

Software today is highly complex, built on APIs, open source libraries and other common code repositories. These pieces make life easier for developers, but when anything needs updating, it can have a rippling impact across a company’s code base. And the larger that code base, the more time it will take to implement these changes, some of which can be essential from a cybersecurity perspective.

Imagine being able to create an automation recipe that cascades across your code making that change automatically in a fraction of the time humans could take to do the same work. That’s what Seattle-based startup Moderne is doing with its code remediation automation platform.

Company CEO Jonathan Schneider, who co-founded the startup with industry veteran Olga Kundzich, points out that almost 80% of code is coming from third-party components. While it’s helping developers avoid reinventing the wheel, it’s also creating massive code bases, which are by the nature of their size harder to update.

That’s where his company comes in, which he says is building a category-defining product. “We’re really the first mover in automating source code remediation itself. So we’re the ‘Do it for me’ of modernization and security, going in and doing source-to-source transformation across big enterprise code bases,” Schneider explained.

He said the reason we haven’t seen other companies try it before in this way is because it’s a very tough problem to solve. “We’re transplanting fixes into somebody else’s code. What makes that really difficult is if the body of code doesn’t recognize that transplanted code as its own, it will reject it. So it’s like organ rejection disease, so the change has to be crafted in a way that looks idiomatically consistent in the context of each little piece of code that we’re inserting it into, and in a large environment like that,” he said.

They create what they call recipes, which define exactly what has to happen for the changed code to fit in the existing code. “It’s a rule-based system that just does a lot of inferencing about the existing style in the code. And then we’re able to craft the change to look like it.”

The founders launched the company in 2020, and built a solution based on some of the work that Schneider had done as engineer working at Netflix. Building on that early work, he and his team were able to build this solution, and he says it turns out that the needs of any company with a complex code base aren’t all that different from Netflix or any other tech-driven company.

Today the company has 20 people, mostly highly experienced engineers. Schneider says these folks tends to be diverse, and that’s helpful as he’s working to build a diverse and inclusive team. “Honestly, perhaps we’re just lucky that it turns out that the population of people that are highly skilled in this way just happen to be more diverse than you would think,” he said.

The company closed a $15 million Series A at the end of 2021, but are announcing it for the first time today. That round was led by Intel Capital with participation from True Ventures, Mango Capital and Allstate Strategic Ventures.

Read Entire Article