Qodo 3.0 adds governance layer for agentic software development
Qodo has released version 3.0 of its AI Code Quality and Governance Platform, targeting the growing challenge of maintaining software standards when AI agents are generating code at speeds that outpace traditional review processes. The update positions Qodo as a governance layer for what the company calls "agentic software factories", where autonomous agents orchestrate work across the full software development lifecycle.
The company cited an industry study in which 91% of engineering leaders said they feared losing control of their codebase, with governance and visibility together accounting for close to 30% of cited risks. Qodo 3.0 is designed to address that anxiety by injecting quality checks at each stage of the agentic pipeline rather than relying solely on post-commit or pull-request-level review.
What is new in version 3.0
The centrepiece of the release is a component the company calls the Wisdom Base, a continuously updated knowledge store that provides agents with awareness of an organisation's coding standards, architectural patterns and pull-request history. Quality controls draw on this store before code is even committed, surfacing standards into an agent's working context and running Qodo's review engine ahead of any pull request being opened.
Once agent-generated pull requests reach the review stage, a PR Triage feature groups related requests into a single work package, giving human reviewers a consolidated view of scope and priority. Reviewers can claim a package to establish ownership and prevent duplicated effort.
A separate tool, the Qodo Software Map, automatically charts an organisation's repositories, models the potential blast radius of a proposed change, and renders quality signals as a heat map across the codebase architecture. This is aimed at engineering and security leads who need architectural-level visibility rather than line-by-line review feedback.
Version 3.0 also extends platform support to Gerrit, adding it alongside GitHub, GitLab, Bitbucket and Azure DevOps. Enterprises with strict data-sovereignty requirements benefit from full on-premises and air-gapped deployment options, and the release adds support for open-source model hosting including Nvidia's Nemotron, so review workloads can remain entirely within a customer's own infrastructure perimeter.
"Every enterprise will build its software factory around its own systems, tools, and ways of working," said Itamar Friedman, chief executive and co-founder of Qodo. "That factory becomes trustworthy when agents can draw on the organisation's accumulated knowledge and operate within its standards."
Market context
Qodo operates in a rapidly crowding segment. Established players such as GitHub (Copilot), JetBrains (AI Assistant) and Sourcegraph (Cody) have moved from code-completion to broader review and governance features, while specialist vendors including Codacy, SonarQube-parent Sonar and Veracode compete on static analysis and policy enforcement. The differentiating pitch Qodo is making is that none of these was designed with agentic workflows as the primary unit of work: most incumbent tools were built around individual developer commits rather than multi-agent task pipelines.
The on-premises and air-gapped deployment capability is a meaningful differentiator for regulated industries. Enterprises subject to financial-services data handling rules, defence sector security frameworks or EU data-residency obligations under GDPR cannot route proprietary code through third-party cloud inference endpoints. Qodo's open-model support is a direct response to that constraint, and it mirrors a broader trend in the enterprise AI tools market toward infrastructure flexibility over hosted convenience.
Funding and trajectory
Founded in 2022, Qodo has raised $120 million to date from a syndicate that includes Qumra Capital, Square Peg, TLV Partners and Susa Ventures, with angel backing from former executives at OpenAI, Meta, Shopify and Snyk. The company did not disclose revenue figures, customer numbers or the valuation at which its most recent funding was raised.
The next milestones investors and enterprise buyers will watch for are reference customers willing to quantify code-quality improvements attributable to the platform, and independent benchmark data comparing Qodo's review accuracy against competing tools in realistic agentic pipeline conditions.