Autoheal AI Inc., an artificial intelligence-native platform engineering startup that’s trying to pioneer the concept of “self-improving software factories,” said today it has raised $7.9 million in seed funding to make that happen.
Today’s round was led by Innovation Endeavors and saw participation from Emergent Ventures, U&I Ventures, Darkmode Ventures, Batch Ventures and Param Hansa Values.
These days, enterprises are shipping more new software than ever before thanks to the widespread adoption of AI tools that can generate code faster than humans. But this acceleration has caused just as many problems as it has solved, with platform engineering teams left to deal with a dramatic increase in production incidents and vulnerabilities in their software, together with rapidly escalating token costs.
To get to grips with these challenges, many teams have embraced a “software factory” model that’s powered by dozens of specialized AI agents to manage each step of the development workflows. However, these agents often fail themselves due to the sheer scale of their rollouts and a lack of shared context and security constraints.
Autoheal says there’s an urgent need for companies to adopt a unified platform for creating, managing and improving these software factory agents, including the all-important coding agents that necessitate their existence. By managing each agent through the same platform, they’ll all have access to the same engineering context, private evaluation infrastructure, cost and security controls.
This is exactly what Autoheal provides, explained co-founder and Chief Executive Sid Choudhury. The startup has built a unified operating model for building, governing and continuously improving AI agents that operate the entire software development lifecycle. Its platform can be hosted with organizations’ private clouds, within a secure boundary, from where it connects to all of the existing coding agents it’s using, along with its code repositories, continuous integration/continuous development pipelines and observability tools.
According to Choudhury, Autoheal builds a shared engineering context graph that’s maintained by two specialized agents: an “Evaluator agent” that rates downstream worker AI agents based on metrics such as CI failures and incident reports, and a “Healer agent” that aims to fix low-scoring agents by opening pull requests that improve model selection, prompts, tools and skills. Each change made by the healer agent is version controlled within a Git, verified against historic benchmarks, and approved by a human supervisor.
Choudhury said he and his co-founders previously spent years building enterprise-grade AI and engineering infrastructure at companies such as Microsoft Corp., ThoughtSpot Inc. and Harness Inc. It was at Harness that they realized the need to “manage agents as code, overseen by continuously learning meta-agents,” he said.
“Our experience taught us that while building the first version of an AI agent is easy, scaling it consistently across the enterprise SDLC is the real challenge,” Choudhury explained. “Platform engineers need a unified platform to deploy agents that don’t just execute tasks, but continuously improve alongside complex enterprise workflows.”
Despite operating under the radar in “stealth” mode until now, Autoheal has enjoyed solid traction with enterprise customers including Normura Holdings Inc., AvidXchange Inc. and Empiric Earth Inc.. These companies all say they’ve used Autoheal’s platform to reduce incident resolution times and save thousands of hours of engineering work.
Normura Bank Chief Information Officer Sameer Jain said his production operations teams were previously overwhelmed by alerts, and had to spend hours endlessly triaging them to manage incidents. Their work often required them to pull engineers away from the things they were working on to help them solve problems. “Autoheal gives us a platform that takes investigation timelines down from hours to minutes,” he said. “The fact it runs entirely within our own cloud, in compliance with our controls, makes it a natural fit for how we operate.”
Autoheal’s goal now is to develop new reinforcement learning techniques that can be used to train customers’ AI agents on their own, private engineering data, Choudhury said. The idea is that customers will be able to create enterprise-specific small language models that operate entirely in their own secure private cloud environments.
Ultimately, it envisages that these SLMs will be used to power each customer’s fleet of software factory agents, reducing costs while enhancing their knowledge of the specific industries in which they operate. Long-term, Choudhury said, he’s convinced that Autoheal’s architecture can be expanded beyond software engineering and into data and security engineering too.
Innovation Endeavor’s Harpinder Singh said he has come across a lot of enterprises that have been asking questions about how they can operate AI agents safely and efficiently at scale, and trust them to run their software factories. “Autoheal is building the agent infrastructure layer that makes that possible,” he stressed. “The opportunity is much larger than one agent or one workflow. It is giving platform teams a repeatable scalable way to deploy specialized intelligence across the engineering organization.”