Salesforce announced a significant shift in its software development process, transitioning to AI agents powered by Anthropic's Claude Code. According to Srinivas Tallapragada, Salesforce's head of engineering, the company has moved its entire development organization to agentic workflows, with Claude Code serving as the primary AI agent. Developers now use unlimited tokens to interact with the AI, marking a fundamental change in how software is built. For April 2026, Salesforce reported a 50.8 percent increase in completed work items per developer and a 79 percent rise in merged pull requests per developer. An ML-based 'Effective Output Score' also improved by 151.3 percent. Despite the increased output, Salesforce's monitoring platform, Engineering 360, showed a five percent drop in incidents. Tallapragada emphasized that quality does not suffer from speed, as safety guardrails and quality standards are integrated into the agentic workflow. "When agentic tools get applied properly, quality doesn't suffer from speed. It benefits from it," he wrote. Salesforce is also developing its own agentic workflows and tools, including 'AI Expert Suite' and 'Salesforce Foundation Plugins,' to support developers. A concrete example cited by Tallapragada was the migration of 33 API endpoints to a new cloud-native architecture, which took 13 days instead of the estimated 231 person-days. "The most important skill today is knowing how to structure problems for an agentic system, when to delegate versus stay in the loop, and how to build reusable patterns your team can compound on," he added. *Source: [thedecoder](https://the-decoder.com/salesforce-claims-ai-agents-cut-a-231-day-migration-to-13-days-with-fewer-incidents/)*