Agents, Autonomy, and the Future of AI in Development Workflows

Developers can now generate and refactor code in seconds with AI, but testing, debugging, and monitoring practices haven’t kept pace. Teams are building faster than they can verify that their software works as intended.

This session explores how agentic workflows can close that gap. Using the Levels of Autonomy framework, we’ll show what AI delivers today, how to build agentic workflows with SmartBear’s MCP Server, and what fully autonomous quality will look like in your CI/CD pipeline in the near future.  

You’ll learn: 

• How to think about autonomy: Map where AI technologies sit on the autonomy curve and what it means for reliable, scalable software delivery. 

• How to use agentic tech today: See real workflows that orchestrate tests, detect flaky runs, and feed actionable insights back into development pipelines. 

• What’s coming next: A first look at autonomous agents that execute, adapt, and collaborate across your CI/CD pipelines, helping teams move at code speed while keeping confidence high. 

If your team is already using AI to write code, this session will show how to extend that intelligence to testing, debugging, and monitoring so you can scale quality alongside velocity and prepare for an autonomous future.