GitLab Transcend Hackathon: What Developers Built on GitLab Orbit

The recent GitLab Transcend Hackathon has shown how developers can transform their workflows when equipped with tools that connect code, deployments, and ownership of every component. With over 1,500 participants and nearly 300 projects submitted, the event made it clear that context—not just automation—is the key to accelerating development without losing control. At Q2BSTUDIO, as a company specialized in artificial intelligence and cloud services AWS/Azure, we saw in this competition a confirmation of what we already apply in our projects: a live graph of relationships between code, pipelines, tests, and teams allows answering in seconds questions that used to require hours of searching across dispersed tools. The winners explored everything from impact analysis in merges to secure migrations with calculated costs. One team managed to reduce CI time by 90% by running only the tests actually affected by a change, thanks to navigating the call graph. Another built an agent that, upon detecting a vulnerability, automatically traces all paths to exposed endpoints. The trend is clear: AI agents need first-hand context, not scattered fragments. At Q2BSTUDIO we apply this philosophy when developing custom software and cybersecurity solutions that integrate data from multiple sources. The ability to orchestrate agents that understand the entire ecosystem—from code to cloud infrastructure—is what makes the difference between fragile automation and solid digital transformation. The hackathon also revealed that agent governance will be the next big challenge: when agents write more code, someone must answer for their actions. Tools like GitLab Orbit, which offer a queryable, real-time updated graph, are the foundation on which to build that trust. At Q2BSTUDIO we help companies implement these capabilities, combining BI/Power BI to monitor performance and cloud AWS/Azure to scale agents. The result: teams that deploy changes with confidence, knowing exactly what they break, which tests run, and who handles each impact.