Day 2 at Dreamforce 2026, and here are the key takeaways shared by our Salesforce Practice Head, Dnyanesh Gawali. As AI agents move from assisting to acting, enterprise architecture has to evolve with them. Trusted data, governance, permissions, and observability are quickly becoming essential. A timely perspective on what enterprises need to prepare for next. #Dreamforce2026 #Salesforce #Agentforce #SalesforceArchitecture #EnterpriseAI #AgenticAI #AressSoftware
While much of the attention on Day 2 at Dreamforce was taken by the Salesforce and Google outage, there was a far more important story unfolding beneath the headlines. The conversation shifted from what AI agents can do to how enterprises should architect for them. A few themes stood out for me: • Agents will need ongoing operations, measurement, and observability, not just implementation. • Business logic, approvals, and controls need to move beyond the UI so humans and agents follow the same rules. • Trusted business context is becoming more valuable than model choice. • CRM may gradually depend less on manual data entry as AI captures and carries context across the revenue cycle. • Identity, permissions, and fraud controls will need to evolve as machine actors begin taking real business actions. The message from Day 2 was clear: AI does not reduce the importance of enterprise architecture. It raises the bar for it. When AI can act, data quality becomes decision quality, permissions become guardrails, and architecture becomes the foundation for safe autonomy. Here are 5 architectural shifts from Dreamforce Day 2 that enterprises should not overlook. #Dreamforce2026 #Salesforce #Agentforce #SalesforceArchitecture #EnterpriseAI #AgenticAI #Data360 #CRM #AressSoftware
-
-
-
-
-
+1