The AI that reads your pull request the way your best senior engineer would - and stops the bad code before it ships.
QODO, FORMERLY CODIUMAI — A NAME BLENDING "QUALITY" AND "CODE." FOUNDED BY ITAMAR FRIEDMAN AND DEDY KREDO, TEL AVIV, 2022.
In the gold rush of AI-assisted programming, most tools compete on one metric: how fast they can produce code. Qodo picked a harder question. As machines write a growing share of the world's software, who verifies it? The company - founded in 2022 as CodiumAI and rebranded to Qodo in 2024 - builds AI agents that review code, generate tests, and enforce an organization's own quality standards across the development lifecycle.
The pitch lands because the bottleneck has moved. When generation gets cheap, review gets expensive. Qodo's agents plug into the tools engineers already use - VS Code and JetBrains IDEs, a command-line agent, and pull-request reviews on GitHub, GitLab, Bitbucket, and Azure DevOps - and apply consistent checks to local edits, pull requests, and CI pipelines alike.
By its March 2026 Series B, Qodo said its enterprise footprint had grown roughly 11x in a year, its tools had reached more than a million developers, and its customer roster had come to include names like Nvidia, Walmart, Red Hat, Intuit, Texas Instruments, and Monday.com.
Intelligence is enough for generation. Wisdom is a must for governance. The future of software is autonomous, but it must also be wise.
Qodo framed its Series B around a phrase that resonates with any engineering leader watching AI velocity climb: software slop - the flood of AI-generated code that looks plausible, passes a glance, and works until it quietly doesn't. The more code AI writes, the larger the pile of changes no human has fully reasoned about.
AI generates code faster than teams can meaningfully inspect it, so risky changes slip through.
Generic tools judge an isolated diff without the codebase history or prior PR decisions that give it meaning.
"Good code" lives in senior engineers' heads and past reviews. Qodo captures those rules and applies them consistently.
An assistant inside VS Code and JetBrains offering code generation, test generation, local review, and codebase-grounded chat.
A hosted, multi-agent reviewer that analyzes pull requests on GitHub, GitLab, Bitbucket, and Azure DevOps - posting summaries, risk analysis, inline comments, and test suggestions. It grew from the open-source PR-Agent.
An agent that generates and validates unit tests to raise coverage and catch regressions before they merge.
A CLI agent for scripting, automation, and running Qodo's agents directly inside CI/CD pipelines.
Qodo describes its newest platform as "artificial wisdom" built from three pillars: a multi-agent fabric of specialized review agents, a context engine drawing on codebase history and past pull-request decisions, and rules-lifecycle management that captures an organization's own coding standards.
The $70M Series B, announced March 2026, was led by Qumra Capital with Maor Ventures, Phoenix Venture Partners, S Ventures, Square Peg, Susa Ventures, TLV Partners, and Vine Ventures - plus angels including OpenAI's Peter Welinder and Meta's Clara Shih. The company said it would use the funds to scale enterprise operations globally and deepen AI-driven governance.
Qodo's model is developer-led: a free tier hooks individual engineers and open-source projects, team plans add compliance and best-practices learning, and enterprise contracts bring analytics, multi-repository awareness, SSO, and air-gapped, on-premises deployment for regulated industries.
| Dimension | Qodo | Generation-first tools |
|---|---|---|
| Primary focus | Verification, testing, governance | Code completion & generation |
| Review scope | GitHub, GitLab, Bitbucket, Azure DevOps | Often GitHub-centric |
| Standards | Learns your org's definition of quality | Generic best practices |
| Deployment | Cloud, on-prem, air-gapped | Mostly cloud |
| Origin | Built around code integrity | Built around autocomplete |
Competitive landscape includes GitHub Copilot, Cursor, Amazon Q Developer, Sourcegraph Cody, and review specialists like CodeRabbit and cubic. Qodo's wedge is the verification layer rather than the generation moment.
Itamar Friedman and Dedy Kredo launch the company in Israel around AI-assisted code integrity and testing.
Raises ~$11M from TLV Partners, Vine Ventures, and angels including OpenAI's Peter Welinder; open-sources PR-Agent.
CodiumAI becomes Qodo - "Quality" + "Code" - alongside a Series A led by Susa Ventures and Square Peg.
Adoption crosses a million developers and lands enterprise customers, including Fortune 100 names.
Raises a $70M Series B led by Qumra Capital ($121M total) and launches its multi-agent Qodo 2.0 platform.
CEO Itamar Friedman is a repeat founder. He previously co-founded Visualead, a visual-QR-code startup acquired by Alibaba, and went on to direct machine vision at Alibaba's Israeli AI lab. His co-founder, Dedy Kredo, previously led product and data-science work at companies including VMware.
Their framing of the market has been consistent: the value in AI coding isn't only in producing more code, but in producing code teams can actually trust. Friedman poses it as a question - "what would it take for agents to judge code the way your best engineers do?"
The name Qodo is a blend of "Quality" and "Code."
Its mascot is an anteater in dark sunglasses inside a purple circle - a nod to sniffing out bugs.
OpenAI's Peter Welinder backed the company before it had even rebranded to Qodo.