Salesforce 发布与 NVIDIA 合作的 CRM 推理模型 Koa，但属产品叙事/访谈，无订单或财务增量

Not every job calls for the same
AI model
. Some tasks, like routing a case or updating an opportunity, need precise multistep reasoning. Others might need to be done particularly quickly or cost-effectively.
So rather than forcing customers to stick with a single tool, Salesforce gives them flexibility to use whichever model works best for any given use case, whether it’s a frontier, open, or custom model like Koa. Salesforce’s
first CRM reasoning model
,
Koa
is purpose-built for the work enterprises do. Built in collaboration with NVIDIA and running inside the Salesforce trust boundary, it’s trained in simulations drawn from 27 years of Salesforce experience running CRM workflows.
It’s orchestrated alongside frontier models rather than in place of them, and that flexibility is central to how Salesforce approaches enterprise AI. As Chief Data Officer Michael Andrew puts it, “We believe in giving our customers choice, that you can pick the model that’s right for your particular use case … and that you have the flexibility to keep evolving and changing what model you use.”
When the Salesforce Newsroom caught up with Andrew, he also explored the flexibility that multiplayer AI now offers users in
Slack
. Instead of one person, one screen, and one agent in a private tab, people and agents work out in the open on shared context so the whole team learns from every exchange. In that setup, Andrew said, the agent is “working with your team together, and it has all the shared context.” No matter what model it’s based on.
Watch the video here, or see the transcript below.
Q: What’s the difference between a deterministic and a probabilistic system?
Andrew:
When you press the power button on your TV, it’s always going to come on. That’s deterministic. You tell your child, ‘Hey, will you turn off the TV?’ Maybe they do. Maybe they don’t.
Agents
are the same way. Software is deterministic. If you open a case about your customer, if a customer files an issue, every time you press that button, you’re going to open a case. Businesses have always run on deterministic systems, which guarantee you’re going to get the same thing every time, everywhere in the world.
You can get a new employee, you can try to train them to do the best possible, but that new employee might make mistakes. That’s what the nondeterministic systems of AI are like — they’re basically big artificial brains. You can train them, you can guide them, but they’re not always going to be reliable.
Q: Why should a company’s deterministic systems be able to connect to any model?
Andrew:
Different models have different strengths and weaknesses, and they’re changing all the time.  All the companies are competing to produce different models, and of course, you have big frontier foundation models produced from partners like Anthropic and OpenAI, but you also have open-source models. You have custom models, like we just released a model partnering with NVIDIA that’s customized from Salesforce.
So we believe in giving our customers choice, that you can pick the model that’s right for your particular use case and cost and value, but knowing that your data is secure, that you can run your process, and you have the flexibility to keep evolving and changing what model you use for the use case. And it’s based on that speed of resolution and cost.
AI is a fuel; it’s more fuel into your engine, and we want to let that fuel come from anywhere, whether it’s Anthropic and Claude, OpenAI and ChatGPT, or any of the many partners that we’re announcing for
AIforce
.
Q: Why is it critical to deliver agentic actions in places where people already work?
Andrew:
We use Slack every day. I think 85,000 employees are talking with their colleagues, talking about their projects, talking about their accounts. You don’t want to have to go somewhere else. Now you can just talk to your agent and say, “Hey, agent, help me out. Suddenly, boom! The agent’s in the chat, and it’s your digital colleague right there in the flow of work with you.
And importantly, in Slack, it also comes with security and permissions. It’s to your channels, to what you want to allow it to do. And I look at Slack’s real power with agents is that it’s teamwork. It’s another team member with you in the same way you work with your team right now. Now you have an agent that’s working with your team together, and it has all the shared context. It can know what you were talking about with that deal or that customer issue. Now the agent has the same context as your team members, and it can now help you with that right context to help you do whatever you need to do.
Go deeper:
Toward
self-improving agents
How we
cut inference spend
by right-sizing our models
Pioneering Salesforce Engineering’s
agentic shift