Whalesbook
AI Code's Trust Problem
AI's rapid code generation is creating a massive amount of software, but the industry struggles to ensure its quality. Despite faster development, trust remains a major issue. Surveys show 95% of developers don't fully trust AI code, and 48% don't always review it before deploying. This means faster code doesn't automatically mean secure software. Qodo aims to fix this gap by focusing on verifying and governing AI-created code, an area many pure AI development tools miss.
How Qodo Builds Trust
Unlike tools that check only small code changes, Qodo analyzes the entire system. Its AI agents examine how modifications affect the whole software architecture, considering company standards, past code history, and risk levels. This smart, context-aware method helps enterprises feel confident when using AI tools like OpenClaw and Claude Code. Founder Itamar Friedman, who previously co-founded Visualead (sold to Alibaba) and worked at Nvidia-acquired Mellanox, was inspired by his experience automating hardware checks. Friedman realized 'generating systems and verifying systems require very different approaches,' a lesson now guiding Qodo to bring 'artificial wisdom'—AI with context and history—to code governance. This approach differs from tools like OpenClaw, criticized for high barriers and security risks, and Claude Code, a helpful assistant but limited with large projects or complex situations.
Trusted by Big Names
Qodo has proven its value, ranking first on Martian's Code Review Bench with a score of 64.3%, far ahead of competitors. This shows it can find complex bugs and issues across files without overwhelming developers. Major clients include NVIDIA, Walmart, Red Hat, Intuit, Texas Instruments, Monday.com, and JFrog. This wide adoption highlights the strong need for reliable AI code verification. These clients are deeply involved in AI: NVIDIA's GPUs power much of its infrastructure, with revenue up 65%. Walmart uses AI in retail. IBM (Red Hat's parent) is growing with AI. Intuit is adding AI agents. Monday.com is also developing AI. JFrog, a DevOps platform, is enhancing AI security. Their AI focus increases the need for secure code, which is Qodo's specialty.
Risks and Challenges
While AI code generation offers great potential, risks remain that Qodo aims to reduce. 'Shadow AI,' where employees use unapproved tools, is common, with over 59% of organizations seeing it. This widespread use is hard to control. Also, 73% of companies say AI coding is too fast for security teams to review, leading to AI bugs in live systems. Around 70% have confirmed or suspect vulnerabilities from AI code. A key worry is that AI might add subtle bugs or security flaws that are hard for tools or humans to spot, especially in large codebases. Qodo addresses the lack of company-specific context, which can cause AI code to miss business rules, compliance, or security policies. As AI is new in critical software, unexpected issues could arise, affecting stability, data, or private information.
Looking Ahead: 'Artificial Wisdom'
Qodo's vision goes beyond just code verification; it looks at how AI will shape software development. Friedman calls this next step a shift 'from basic AI to smart systems—from intelligence to 'artificial wisdom.'' This means AI that deeply understands context, rules, and past actions, acting wisely. As companies rely more on AI for software, the need for tools that ensure code is trustworthy and high-quality will only grow. With its significant funding and strong client base, Qodo is set to lead this crucial development.