Salesforce Inc. today launched Agentforce Operations, an artificial intelligence system designed to extend specialized AI agents into the back office to automate manual work. 

Workflow automation tools have existed for a long time — not because office workers are lazy, but because it’s more efficient to have a task handled by a machine. The problem is that it typically takes an engineer or information technology worker to develop, configure and maintain that automation. That also means someone needs to watch it forever because someday something will change, and it will break. 

The new product takes this concept and puts it in the hands of intelligent machines: AI agents. 

“Agentforce Operations is us tackling the back office,” Senior Vice President of Product Marketing Sanjna Parulekar told SiliconANGLE in an interview. She described the target as “the boring stuff that is a complete time suck” — the manual, repetitive work that often clogs up business operations. 

From the front office to the back office 

Specialized agents that understand company culture, business logic, security, workforce roles and company data take control to build the automations. They also maintain them over time. That means if something changes on the back end, they can adapt. It also means they can remain compliant with security audits and regulatory needs, while keeping employees in the loop whenever critical changes happen during everyday tasks. 

If a change needs to happen, for example if a new regulation drops, a manager can update the configuration using natural language in an email. The agent will work through the update, provide a plan and seek approval. Once locked in, it will implement the change. 

The company said the entire engine operates on “radical transparency,” meaning every action the AI takes is recorded to an audit trail, allowing IT teams to guide the agents, find out where things went wrong and quickly get back on track. 

The launch also builds on Salesforce’s acquisition of Regrello, a company focused on automating business processes, particularly in supply chain operations.  

Parulekar said Agentforce Operations is “sort of the relaunch of that for more industries than they had in the past.” She added that Regrello’s work helped Salesforce think through the deeper pain points of process automation: how to let business users upload a process and get immediate optimization, while still preserving determinism and human oversight where it matters. 

That balance is important because enterprise automation cannot simply be handed over to AI without guardrails. As Parulekar put it, “you don’t want AI just taking action willy nilly out here.” The point is not to remove people from the process entirely, but to give them more control over where humans need to stay involved and where agents can take over the repetitive work. 

Salesforce added that with Operations, the company is providing blueprints for common workflows. There’s no need to start from scratch; it allows users to get going within minutes, or what the company said is 80 times faster than legacy providers. 

“Salesforce’s introduction of Agentforce Operations marks an important step forward in bringing AI-driven automation to the back office,” said Ian Kahn, principal, commercial and service excellence platform leader at PwC US. “Together, we’re helping clients use Agentforce to transform manual operations into intelligent workflows.” 

Salesforce said Operations is available today. Broader ecosystem integration, including the ability to auto-synchronize data and trigger actions with Salesforce Flow, the company’s no-code automation tool, is expected to enter beta mode during May 2026. 

 Image: SiliconANGLE/Microsoft Designer