Prakhar Dixit, Tariq Shaukat and Ali-Reza Adl-Tabatabai discuss why widespread experimentation with AI agents has not consistently translated into scaled business impact. They distinguish isolated coding assistance from redesigning the software development lifecycle, with changes to workflows, tooling and the responsibilities of the people using them.
Prakhar Dixit, Tariq Shaukat and Ali-Reza Adl-Tabatabai compare metrics for small teams and large engineering organisations, emphasising throughput, quality and developer experience rather than token consumption alone. They explain how implicit organisational knowledge must become usable context, and how incremental automation of dependable tasks can build trust in more autonomous agents.
Watch on YouTube
