Salesforce is advancing its agentic enterprise initiative as the next phase of business technology, focusing on embedding AI agents directly into workflows instead of treating them as separate tools. This strategy aligns with enterprises seeking to move beyond isolated AI applications and integrate them into daily operations for employees and customers. The approach raises important considerations about how agents interact with data, business logic, and governance systems while maintaining existing processes.
The company’s latest updates to Agentforce, Data 360, and its broader platform show this direction. Developers will particularly benefit from reduced fragmentation, as disconnected tools have historically hindered productivity. Research from theCUBE Research in 2026 showed tool fragmentation declined from 74% in 2024 to 61% this year, indicating progress toward unified environments. Paul Nashawaty, principal analyst at theCUBE, observed that developers now favor integrated workflows over fragmented solutions.
Salesforce is meeting this demand through APIs, the Model Context Protocol, and a command-line interface. These tools enable agents to access Salesforce capabilities within users’ existing environments, preserving trusted data and governance. Data 360 supplies the necessary business context for agents, while Customer 360 provides workflows agents can utilize without requiring enterprises to rebuild core systems.
A major emphasis lies in balancing automation with human oversight. Agentforce Operations allows agents to manage repetitive back-office tasks, adjust workflows dynamically, and maintain audit trails while keeping human control over critical decisions. This balanced method aims to simplify implementation, as demonstrated by Help Agent, a preconfigured solution that connects to company knowledge bases, actions, and communication channels. Pricing for this tool shifts from usage-based models to outcome-based metrics, aligning costs with completed customer-service resolutions, according to Kishan Chetan, executive vice president of Agentforce Service at Salesforce.
Chetan told theCUBE that integration does not require migration. The system builds on Salesforce’s existing platform to streamline adoption.
theCUBE, SiliconANGLE Media’s livestreaming studio, will examine these developments during its Dreamforce coverage on September 25. The event will feature interviews exploring Salesforce’s agentic enterprise strategy, challenges in scaling AI agents in production, and expected business outcomes. Topics will cover adoption trends, data governance, and human-AI collaboration.
Viewers can access theCUBE’s coverage through its website and YouTube channel, both live and on-demand. Additional discussions appear on SiliconANGLE’s “theCUBE Pod”, a podcast hosted by John Furrier and Dave Vellante, which covers enterprise tech trends including AI, cloud computing, and workplace culture. The weekly “Breaking Analysis” program also provides insights by combining theCUBE’s research with spending data from Enterprise Technology Research.
Salesforce executives, customers, and partners will participate in these discussions, offering direct insights into Agentforce adoption, data trust, and governance. The focus remains on practical implementation—how businesses can integrate AI agents without disrupting existing systems or losing control over key processes.
TheCUBE’s analysis will also assess whether Salesforce’s strategy can resolve ongoing AI adoption challenges, including data silos, permission conflicts, and risks of automation replacing human judgment.