Code AI forgotten: logcat.ai raises $2.55M to put agents to work on device apps

The past two years have changed the world of software development, but there is at least one area that remains untouched by artificial intelligence: the operating system layer inside phones, cars, and other connected devices.

A Seattle startup called logcat.ai has raised $2.55 million to change that.

Co-founded by CEO Varun Chitre and CTO Tarun Vashisth, two engineers with years of experience building device software, logcat.ai builds a system of AI agents that automatically hunt for bugs in every kernel, modem, and firmware of devices running Android or Linux.

The pre-seed round was led by Founders’ Co-op, with participation from Act One Ventures, TheFounderVC, Shorewind Capital, Clayoquot Capital, and Alumni Ventures.

“It’s one of the most difficult areas of application engineering, and it doesn’t explain much. Operating system engineering is hidden today,” Chitre said in an interview.

It is also a challenge for many companies given the lack of engineers working in this field, compared to the much larger number of engineers who create applications and software that runs on top of the operating system.

How does this work: A developer using logcat.ai uploads log files generated by a device when something goes wrong — such as bug reports and kernel logs — and the logcat.ai software analyzes them together to find the root cause and pinpoint where in the code to fix it. Each find cites the exact entry line it comes from, so the developer can check the work.

In the meantime, logcat.ai finds the root cause and recommends a fix. The big plan is for the AI ​​to write fixes, test them, and eventually build new features on its own, with developers approving the work before it’s deployed.

The long-term goal, Chitre said, is to become a standard tool for building and maintaining applications on new and existing hardware — from smartphones to cars to robots and other embedded systems — so the company can outsource full-stack expertise to employees.

“We are looking at a world where software and intelligence go beyond laptops and phones, yet the tools to build high-quality products in that world are still lacking,” said Aviel Ginzburg, general partner of Founders’ Co-op, in a statement.

He called Chitre and Vashisth “one of the only teams in the world to succeed in this challenge.”

Pulling: The company says it has served hundreds of engineering teams in public beta, analyzed more than 10 billion lines of trace, and performed thousands of automated investigations. It is making money but is not yet ready to disclose numbers or customers.

Competitive landscape: Chitre said logcat.ai’s biggest competition is not the other product but the internal documentation and knowledge locked away in the heads of a few senior engineers. Application-level crash tools such as Google Crashlytics and Sentry stop at the application layer and do not perform deep system debugging.

Specialist vendors and contract manufacturers who build equipment can be more potential allies than competitors, Chitre said, as they face similar engineering shortages.

GeekWire first reported on logcat.ai in March, in the Startup Radar roundup.

Team: Chitre and Vashisth met at Esper, a Bellevue, Wash.-based device-management company, where they worked together for more than seven years. They started logcat.ai because they had spent years manually debugging and knowing what was missing.

Chitre has spent over 13 years in the field, getting operating systems up and running on new hardware and porting new releases of Android and Linux kernels to older devices. He was also the maintainer of LineageOS, a widely used open source version of Android.

Vashisth has led development teams working across Android, Linux, and iOS, and brings background to large distributed systems. At Esper, he rose to senior manager of software engineering. His prior experience includes architecture platform engineering at Target.

Currently, the company has only two founders: Chitre in the Seattle area, Vashisth in Bengaluru, India. They plan to hire about 10 people next year, with a distributed team working remotely wherever they can find special talent.

They know that those jobs will not be easy to find, given the shortage of people in this sector. “That’s the same lack our existing product has to deal with,” Chitre said, “and we’re not comfortable with it.”