Tessl raises $125M in $500M+ funding to build AI that writes and stores code

Many startups and large tech companies have taken the initiative to build artificial intelligence into code software. Now, another new player is stepping out of the shadows to throw his hat in the ring, with the goal of fixing the many problems that will arise when humans and all those AIs are all coding together.

Tessl is building what it describes as a “native AI” platform that developers and their teams can use to build and maintain software, and on Thursday it opened a waiting list for those interested in trying it.

We say “under construction” for a good reason: The Tessl product has not yet been launched, and the plan is to be ready early next year. But the London-based startup is now sharing a little more about what it’s doing with some cash in mind: Tessl has quietly raised $125 million in a seed round and a Series A, both announced for the first time today. The latest round was led by Index Ventures, with Accel, GV, and Boldstart participating. GV (aka Google Ventures) and Boldstart co-led the seed round.

TechCrunch has confirmed through multiple sources that Tessl’s post-funding valuation is north of $500 million.

As you can imagine, one reason why a company with no customers or a shipped product gets this kind of attention from top-shelf VCs is because of who is building it.

Tessl’s CEO and founder is Guy Podjarny, an engineer of sorts. His last startup was Snyk, a cybersecurity company that was last valued (by 2022) at $7.4 billion. Prior to that, he was CTO of Akamai, a role he assumed after Akamai acquired his first startup, Blaze, which focused on speeding up website load times.

“Podjarny has an amazing vision and he thinks about his business,” said Carlos Gonzalez-Cadenas, a partner at Index who leads investments. “He is very good [at understanding] engineering communities and building engineering-centric businesses.”

Podjarny said in an interview that the idea for Tessl came from what he did at Snyk.

Snyk focuses on finding (and fixing) security vulnerabilities in code, and Podjarny noted that the same problem is becoming more urgent in code and software interactions in general – especially due to the rapid increase in code written automatically by AIs.

“What is AI doing in software development?” he remembered wondering. The answer was: to speed up, but also to create more automation. The process of maintaining and deploying updates to that code can add complexity and potential for system crashes. This ends up having many negative consequences (safety, overtime, costs, efficiency) for organizations. “When that image formed in my mind, that’s when I knew I was going to build it,” he said.

The startup’s name, Tessl, refers to “tessellation,” Podjarny said, as it aims to make sure that software and the code behind it fit together, rather than existing in a messy, overlapping jumble.

Podjarny was concerned about what kinds of applications or code he thought would be built, or maintained, on Tessl. But it sounds like it’s actually going to start slowly.

“We haven’t shared a complete strategy yet on what that is,” he said of the targeted applications or use cases. “I can say that we don’t start with games. We start with simple software that allows us to build an end-to-end system that is more manageable for LLMs to implement, and more manageable for people to specify. And we will come from there. ”

The basic idea behind the implementation goes something like this: Developers and their teams (including product managers and others who don’t write code) can provide specifications to Tessl in the form of natural language or code. Tessl will then write code that matches those features.

Teams can test that code in a sandbox, where issues can be flagged and resolved, and continue to tweak the details as needed. After that, Tessl can be automatically configured to store that code in that definition. So if something else is at risk of being broken due to new code, Tessl will use a fix to identify and fix it.

It sounds like Tessl is not considered a walled garden. Podjarny said he is talking to others who have built or are developing AI coding assistants, with the idea that these other platforms will also be powered by Tessl. The startup will aim to support Java, Javascript and Python, and will add more languages ​​over time, Podjarny added.

One of the reasons why investors like this idea and support it is because of that expansion. Preserving the code is something that has been “highly signaled” as important right now, Gonzalez-Cadenas said. “But you’re building a recording system here,” he added. “Once you do that, there are a variety of opportunities.”