AI startup Autoheal raises $7.9 million to build self-improving software factory for enterprises

AI startup Autoheal has raised $7.9 million in seed funding to expand its platform for deploying, governing and improving AI agents across software engineering workflows.

Advertisment

The round was led by Innovation Endeavors, with participation from Emergent Ventures, U&I Ventures, Darkmode Ventures, Batch Ventures’ CTO Fund and Param Hansa Values. Harpinder Singh of Innovation Endeavors will join Autoheal’s board.

The round also included angel investors Shawn Kung, Sumeet Arora, Anshu Sharma, Savin Goyal and Srikant Gokulnatha.

Following the fundraise, Autoheal plans to build enterprise-specific small language models trained using reinforcement learning on each company’s private software development lifecycle data. It also expects these models to handle some tasks inside its software factory that do not require generative AI.

According to Autoheal, the growing use of AI coding tools is creating additional operational work for engineering teams, including production incident response, security vulnerability remediation and managing AI-model costs.

Its platform, which it describes as a “self-improving software factory”, connects coding agents with code repositories, CI/CD pipelines, observability tools, cloud runtimes and issue trackers.

These systems feed into an engineering “context graph” that gives agents shared information about an organisation’s software environment.

Autoheal also uses two specialised agents called the Evaluator and the Healer. The Evaluator scores worker-agent performance using signals such as pull requests, review comments, CI failures and production incidents. The Healer can then propose changes to prompts, tools, skills or model selection.

Changes are tested against historical runs, version-controlled in Git and require engineer approval before deployment.

Sid Choudhury, co-founder and CEO of Autoheal, said scaling AI agents consistently across the enterprise SDLC is the bigger challenge.

“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,” he said.

Autoheal’s technology is being used by Nomura Bank and AvidXchange for engineering workflows including alert triage and production incident response.

Sameer Jain, CIO, Wholesale at Nomura Bank, reported that Autoheal had reduced investigation timelines “from hours to minutes” while operating within the bank’s own cloud and controls.

The startup grew out of its founders’ experience at Harness, Microsoft Azure, ThoughtSpot and AppDynamics.

Over the longer term, Autoheal plans to extend its architecture beyond software engineering into data and security engineering.