The Agentforce is here. Salesforce wrapped one other edition of its annual Dreamforce conference this September. Joining the swarms of attendees — and the swarms of Waymos shuttling them around an extra-cleaned San Francisco — we each now have a swarm of agents at our fingertips to remodel work, neatly controlled inside the Salesforce ecosystem. While Dreamforce is all the time a spectacle for its marketing-honed pronouncements of the long run, this 12 months provided an unexpectedly compelling vision of how AI-based agents are about to revolutionize the workplace and customer experience.
Let’s temper our expectations for only a bit longer. Benioff pondered at his keynote, “Why would our agents be so low-hallucinogenic?” Yes, they’ve the information, the metadata, the workflows and an enormous array of services to attach into; and as long as your systems only live inside Salesforce, it sounds pretty ideal. Salesforce may or may not have invented prompt engineering, a claim Benioff also made within the keynote evoking perhaps the “Austin Powers” Dr. Evil monologue about his father inventing the query mark. But can Salesforce meet the Agentforce vision? In the event that they do, it’s going to be a giant deal for the way work gets done.
Let’s be real though: our systems and data don’t all live inside Salesforce. If the long run of labor is defined by groups of agents working together, how far can walled gardens and closed ecosystems really get us in delivering outcomes across our businesses? Actually Apple and Microsoft and Amazon and a number of others wish to wrap their arms around the large agent opportunity in front of us. But as each wave of technical advancement has brought forward different flavors of closed vs. open debates, we’re ultimately going to want a regular for agents working along with each other across boundaries. Otherwise only parts of your online business will meet this chance.
As we frequently do when faced with the open/closed conundrum, allow us to look to the open web as a way forward. Similar to apps in your phone need an online view to enable an infinite array of mobile app outcomes, the identical might be needed within the upcoming multi-agent frontier. Tools like Slack provide UI frameworks like Block Kit that may power the user interface for a straightforward agent interaction, however it’s not cut out to handle the depth of recent user experiences. Take Clockwise Prism for instance. We’ve built a next-level scheduling agent for locating time for a gathering even when there isn’t any current “whitespace” on the calendar tomorrow. When hooking it as much as other agents to land that unattainable meeting together with your hottest sales prospects, you’ll need a solution to either confirm or explore a myriad of sophisticated and powerful scheduling options. Providing an online view for doing that is the clear path forward.
Throughout his keynote, Benioff repeated the mantra that you just don’t want DIY agents inside your online business. And he’s right. Enterprises want controlled and simplified workflows that deliver repeatable value. And yet they don’t wish to be stuck in a silo. This is the reason we want an open standard for the multi-agent future. We’d like a dependable way for agents to interact with each other, to cross boundaries across applications and ecosystems and do it in a way that keeps businesses accountable for their product experience.
You may be just as more likely to kick off a set of labor agents from inside an Atlassian Jira ticket connected to a Salesforce customer case as you’ll wish to kick off a set of agents in reverse originating from inside Salesforce connected to Atlassian. For agents to work together no matter where a piece request originates and in any variety of directions with a consistent user experience, again a regular for doing this is required.
What else needs to be represented on this standard? Outside of Salesforce, the multi-agent ecosystem today is an exciting wild west. On the every day we see latest innovations and ways of connecting and constructing AI systems and agentive workflows. One recent tie-in between the AI framework LangChain and a tool called Assitant-UI brought this insight:
Indeed, we’ve already covered how crucial user experience is to agents. And clearly agents must give you the chance to quickly stream their responses when working with other agents. But what of generative UI and human-in-the-loop of their application?
Let’s start with human-in-the-loop; one other area of broad agreement. While Salesforce and others talk a giant game about automation, it’s all the time grounded in the necessity to give you the chance to bring a human back into the middle when crucial. We learned this lesson as well at Clockwise, and have built our scheduling agent experience around a core concept of with the ability to check back in with the user with a proposed set of scheduling options. When you find yourself doing complex work, it’s amazing to get to full automation, however it starts on the backbone of involving the user and keeping them within the loop. Any standard have to be built around an optional ability to check-in and make sure with the user before proceeding, and eventually allow for full automation when confidence is high enough.
And so what of generative UI? Here I’d propose that what is required isn’t necessarily generative UI but “native UI.” What is essential is that the agent is producing a UI that’s native and controlled by the service/agent responding to the request. Only the native service could have the context and understanding crucial to render a user interface that ties into the agent request. Whether that UI is rendered using generative AI or another non-AI mechanism is left to the responding service as an implementation detail. And so here, we expect the open standard must allow for the responding service to regulate and deliver native UI to an agent request.
What comes next? We’re excited to proceed to look at what an open multi-agent future might appear like. We’ve created a draft of something we’re calling Open Multi-Agent Protocol (OMAP) and we’re excited to maintain pushing the conversation forward. It won’t be long before there are entirely latest kinds of jobs on the market where people use agents to do work in powerful and streamlined ways. The age of the Agent Orchestrator job description is upon us, and while Salesforce paints a compelling path ahead, we’ll need a regular way for agents to interconnect beyond boundaries.
