Sigma Software helped Infobric, a leading construction site software provider, scale Claude adoption across engineering, operations, and business functions, matching AI autonomy to the needs of each team and workflow.
Infobric was evolving its real-time construction access management platform and wanted to increase engineering velocity while reducing manual efforts across a broader set of organizational functions. As Infobric’s long-term product development partner, we were engaged to establish a scalable AI adoption model that could support different levels of AI automation while maintaining oversight of business-critical decisions.
Using Claude Code and Claude Cowork as the foundation, we helped the Client establish, implement, and scale an AI operating model that supports both human-led and agent-assisted work organization-wide. The Client achieved 30-40% efficiency gains in selected engineering processes, increased release cadence to nearly four releases per quarter, and accelerated activities ranging from troubleshooting to design preparation.
Infobric is a leading provider of construction technology across the Nordics, developing security and access control solutions for construction sites. Operating multiple parallel development streams with varying levels of complexity and maturity, the Client wanted a practical framework that would help teams benefit from automation without compromising quality, accountability, or engineering standards.
Rather than pursuing a one-size-fits-all approach to AI adoption, we helped Infobric establish an operating model that aligned the level of AI autonomy with the function and task. Some activities required a human tightly in the loop, others benefited from human-led AI augmentation, while well-defined and repetitive work could be delegated to dedicated task-specific agents. This approach allowed teams to adopt AI at different speeds while operating within a common framework.
As teams became more comfortable incorporating Claude into their daily work, its role gradually expanded beyond software development. We helped the Client apply the same capabilities to operational challenges where engineers were spending considerable time investigating incidents, reviewing infrastructure changes, and documenting decisions. As a result, Claude became part of day-to-day DevOps and incident-management activities, enabling:
Dmytro Gorlov
Head of Information Systems
Infobric
When we recognized how powerful this technology was, we didn't chase adoption everywhere at once. We started where we were confident AI would amplify real value, not amplify activity for its own sake. Claude has consistently proven to be the most efficient model on that path.
