Introducing the Context Quality Score (CQS): a new framework for measuring and optimizing AI context
Atlassian introduces the Context Quality Score (CQS) to measure and improve the context feeding AI models, addressing widespread enterprise AI pilot failures.
Atlassian highlights the persistent challenge of connecting relevant context to AI workflows, noting that effective AI performance depends more on context than model sophistication. To address this, the company developed the Context Quality Score (CQS), a metric designed to evaluate enterprise data across three dimensions: completeness, connectedness, and relevance. Each dimension targets a specific data blind spot, providing teams with actionable insights to enhance the quality of information fed into AI systems. The framework aims to bridge gaps between data and AI outputs, ensuring more accurate and efficient results.
The CQS framework is built on three core pillars validated by research to improve AI output quality while reducing token usage per run. When integrated with the Teamwork Graph, teams can engineer and activate seven types of context stored within the graph, enabling more reliable AI-driven processes. The approach emphasizes practical application, offering a roadmap for teams to identify and address context gaps across the software development lifecycle (SDLC).
A whitepaper accompanying the announcement details how CQS assesses context health and demonstrates the Teamwork Graph’s role in powering a context-aware SDLC. The document provides actionable strategies for teams to enhance their AI tools by ensuring connected data is available for building, testing, and deployment. The focus is on delivering measurable improvements in AI performance through better data integration and retrieval.
Enterprise frustration with AI investments persists despite high spending, with a 2025 MIT study revealing 95% of AI pilots failed to impact organizational bottom lines. Atlassian attributes this to insufficient context in enterprise data, rather than model limitations. The introduction of CQS and its integration into the Atlassian Ecosystem aims to address this gap by enabling enterprise-grade AI integrations through the Teamwork Graph.