May 2026 Summaries

The evolution of code review processes has become increasingly necessary as AI-generated code contributes significantly to pull requests, prompting the release of Qodo platform v2.3 with a new Findings Page to address this shift. This update provides engineering managers and leaders with a consolidated view of issues across repositories, allowing them to gain high-level insights into risks and trends without manually compiling data. The Findings Page, available in beta for various Git platforms, offers a comprehensive overview of critical findings, resolution rates, and average issues per pull request. It includes filters to focus on specific services, contributors, or issue types, thereby facilitating risk-based triage and trend awareness. As AI continues to scale code production, the Qodo platform aims to transform code review into a more strategic, portfolio-level discipline, enabling leaders to efficiently manage and act on review data for improved codebase health.

Qodo's code review system evolved from a simple, single-command prompt to a sophisticated multi-agent architecture, presenting a challenge in maintaining accurate benchmarks. Initially, the system utilized a single LLM call to return code suggestions in a YAML format, which was straightforward to measure. However, as the system expanded into a multi-agent pipeline incorporating specialized agents for context collection, issue finding, and compliance enforcement, the original benchmarking method became inadequate. The new architecture required a shift in evaluation strategies to account for the complexity and non-determinism of the multi-agent system. This led to the development of a new benchmarking infrastructure using synthetic pull requests and LLM-as-Judge with ensemble voting to ensure precise evaluation of agent performance. By focusing on precision and recall across agents and utilizing ensemble judges, Qodo improved its ability to diagnose and address system failures, transforming the evaluation process from a static leaderboard metric to a dynamic, interpretable feedback loop. This methodological shift not only enhances the system's reliability but also provides a framework for other teams to evaluate multi-agent systems effectively.

HiBob, a comprehensive HR platform, utilizes AWS infrastructure to support its global user base and incorporates AI tools to enhance its development lifecycle. While AI tools like Cursor, Claude, and Notion initially boosted productivity by quickly turning product requirement documents into code, they also shifted the bottleneck from coding to code review. To maintain high code standards without slowing down development, HiBob adopted Qodo as a mandatory AI code reviewer, which reads code changes against the full codebase and business requirements, catching issues human reviewers might miss. This decision was reinforced after a critical bug slipped through traditional reviews but was identified by Qodo, prompting HiBob to integrate Qodo fully into its software development lifecycle. Now, the review process involves a "double-check" system where developers must address Qodo's comments before human reviewers finalize the pull request, effectively balancing automation with human oversight and ensuring high-quality standards are consistently met.