Martin Harrysson and Natasha Maniar from McKinsey's Software X discuss the paradigm shift occurring in software development due to AI, focusing on the people and operating model aspects. They highlight a disconnect between the potential of AI and the reality of only 5-15% productivity improvements in most companies due to bottlenecks in collaboration, code review, and tech debt. They suggest moving towards AI-native workflows and roles, such as smaller, spec-driven teams and consolidated engineering roles. They share a client study where interventions like using agents for sprint story assignments and reorganizing squads by workflow led to significant increases in delivery speed and code mergers. The speakers emphasize the importance of change management, upskilling, and a robust measurement system to scale AI adoption effectively across large organizations.
Sign in to continue reading, translating and more.
Continue