Cresco Enhances Trust Code Hub with Multi-Perspective AI Review Feature OmniReview for Software Quality

Cresco Enhances its Software Quality Solution with OmniReview

Cresco, a leading IT service provider based in Tokyo, has recently announced an upgrade to its software quality solution, Trust Code Hub, by adding a new feature called OmniReview. This innovative multi-perspective AI review capability aims to significantly enhance the quality and efficiency of software development processes.

The Trust Code Hub already specializes in analyzing the quality and vulnerabilities of source code. However, with the addition of OmniReview, it expands its scope to encompass reviews of upstream documents such as Requests for Proposals (RFPs), requirement definitions, and basic design documents. This is a significant step towards improving both the quality and productivity of software development.

Addressing System Development Delays

According to the 2016 Software Metrics Survey conducted by the Japan Information System Users Association, over 50% of project delays in system development originate from issues in the requirement definition phase. The most prevalent factor contributing to these delays is the inadequacy of requirement specifications. Oversights made during the upstream process can lead to catastrophic errors downstream, a problem that intensifies with the accelerated development pace enabled by generative AI.

Even as code generation speeds up, if the speed and accuracy of reviews do not improve correspondingly, quality assurance could become the bottleneck of the entire development process. Consequently, establishing an effective review system upstream proves to be a challenging task. Professionals like project managers, architects, and UX experts are often too busy to align their schedules for comprehensive reviews, resulting in a tendency to settle for superficial compliance updates. Moreover, reviewers frequently focus solely on their areas of expertise, potentially overlooking critical perspectives from other domains. As long as quality assurance relies on individual experience, delivering comprehensive quality remains a daunting challenge.

The OmniReview Feature

The newly introduced OmniReview feature within Trust Code Hub automates the design of necessary review perspectives and expert roles by analyzing the content of submitted documents. This document review support system leverages multiple AI agents to conduct multi-faceted reviews, allowing for consistent review quality across various projects, irrespective of personnel expertise or tight deadlines.

Key Features of OmniReview

1. Automatic Review Team Formation – By analyzing the document’s content, OmniReview automatically generates the necessary reviewer roles. The system can assemble a multidisciplinary team including quality control, legal, security, and UX experts tailored to the document's specifics. Users can also incorporate their designated reviewers into the system.
2. Multi-Faceted Parallel Reviews – Reviews are conducted in parallel according to different roles, allowing for the consolidation of unique insights while eliminating redundancy. Reviews can be scored from 0 to 5, simplifying the prioritization of revisions.
3. Accumulation of Organization-Specific Perspectives – Organizations can now save and reuse their internal standards, regulations, and checklists as reference material, transforming individual experiences into organizational assets for review perspectives.
4. Compatibility with Various Document Formats – OmniReview supports multiple document formats including PDF, DOCX, PPTX, and XLSX, accommodating a wide range of materials from requirement definitions to operation manuals and academic papers.
5. Human Final Judgment – While OmniReview efficiently covers comprehensive review perspectives and generates initial feedback, the final judgment on whether to adopt suggestions and assess overall quality rests with human reviewers.