You Don't Need a Bigger AI Budget at Dreamforce. You Need a Roadmap.
I'm heading to Dreamforce this week (September 15 to 17, at Moscone Center in San Francisco), and the theme will be everywhere before you even badge in: The Agentic Enterprise. Agentforce will, of course, get its own keynote. But, so will Claudeforce, which is Salesforce's expanded partnership with Anthropic. And now, just days before the show, Salesforce is previewing the next piece of its "AIforce" strategy: a Trusted Enterprise AI Harness and a companion AI Control Plane, built to make agents trustworthy enough to act, not just suggest. Data 360, the platform formerly known as Data Cloud, gets top billing as well.
Is An Agentic Enterprise All or Nothing? No.
For most industries, the continued growth of Salesforce agentic support is an exciting pitch. But, if you're a bank, credit union, or fintech, there's a trap I'd like to gently steer you away from: treating this as an all-or-nothing proposition. Either you have to launch a full agentic transformation the moment you're back from San Francisco, or you hold off on Agentforce entirely until every system is perfectly unified. It really doesn't have to be one extreme or the other and in my opinion, it shouldn't be.
The challenge for most folks is wrangling the data for AI. Your customer data doesn't live in one place. It's scattered across a core banking system, a loan origination platform, a KYC or AML vendor, maybe a payments API or a banking-as-a-service partner if you're a fintech, with Salesforce sitting somewhere in the middle trying to be the system of engagement. In the midst of all of that, an AI agent will only be as good as what it can see. Point it at a customer record missing half the relationship, half the transaction history, half the risk signal, and AI will fail with confidence! Your AI agent has a conversation with a customer, then gets the balance wrong or routes a fraud flag to the wrong queue. This could do real harm to customer trust and could even possibly trigger a regulatory conversation.
That risk is In the specific use case, not the AI agent, itself, so start safe and build out from there.
Phase 1a: Quick Wins
If you're data is the challenge, that's OK. It's possible to avoid the headier AI tasks and still harvest huge value from AI. One of the safest places to begin that generates real lift is to look at the Agentforce for Financial Services functionality ships with out-of-the-box templates, including Banking Customer Service Assistance, Financial Advisor Assistance, and Insurance Service Assistance. These are all grounded natively in the Financial Services Cloud data model and protected by the Agentforce Trust Layer. Salesforce describes them as ready to deploy out of the box, with minimal lift, unlike generic AI tools that need months of configuration. You don't need enterprise-wide identity resolution to stand these up. You need to turn them on.
With those wins under your belt, you can start building out your enterprise agentic trust layer.
Phase 1b: Enterprise Governance
Here's where Salesforce's new AI Harness and Control Plane actually matter to you, not just to your architects. Think about a routine question like whether a customer qualifies for a fee waiver. Right now, no single system has the full answer. Your CRM knows the relationship. Your core banking platform knows the account status. Your compliance team knows the policy. A harness is what lets an agent pull all of that together and act within your rules instead of guessing. Remember, this is not confined to your CRM. This is an enterprise solution.
Whereas a harness orchestrates agentic decisioning, a control plane is lets you leadership team see which agents are running, what they're doing, and what they cost, across the enterprise. This means you have now have a way to measure your running ROI. You don't need to understand the architecture to appreciate the outcome. You now have a clear view of what you're getting for what you're spending. No surprises.
Phase 2: Unified Customer Identity and Activity
That visibility and governance is exactly what makes Phase 1 your funding mechanism for Phase 2. Every hour of contact center volume a service template absorbs, every meeting prep a Financial Advisor Assistance agent shortens, helps you build the business case, and often the actual budget, for the harder work: resolving customer identity across core banking, KYC, and payments data inside Data 360. That's the unglamorous phase nobody puts on a keynote slide. It's also the one that determines whether your next wave of agents is safe to trust.
Phase 3: Enterprise Trusted Data for AI Agents and Advisors
While Phase 1 and 2 have improved the bottom line, this is the big payoff. Once identity is resolved, the same platform supports agents that need the full picture, an advisor assistant assembling complete household and portfolio context before a client meeting, or a service agent working a case end to end. Those work because the data underneath them finally deserves the trust you're placing in it.
Expect Your Roadmap to Be Different
No two companies are alike. While we see a lot of similarities across institutions, it's important to realize that any roadmap needs to reflect your challenges and where you'd like to end up, when all is said and done:
You're starting from your own unique state of the enterprise, including your technology and your resources.
You have different process technical challenges and your pain points are unique
You have different business goals and timelines
Those goals and timelines may change and evolve
Your board wants measurable ROI, not an experiment
Ideally, you want to see immediate returns on the initial efforts, to solidify executive buy-in
This is why it's important to always align the technology to the business roadmap and not the other way around.
Final Word
If you're walking into Dreamforce this week, don't just ask what Agentforce can do. Ask what you could deploy soon that starts paying for what you'll need next. You don't need a bigger budget to start. You need a roadmap. Plan to sketch the first phase of it this week.
If you're not sure what business value can be captured or what AI options make the most sense for your vision, no worries! You bring the expertise of your business. We'll bring our expertise in Salesforce and AI. Together, we can build the exact roadmap that delivers immediate value and long term success for your business!
Learn more:
DX Foundation helps you leverage AI and advanced data solutions to enhance Agentforce Marketing, CRM, and Agentforce Service.
Our unique expertise enables faster, smarter implementations that drive ROI and empower agents to deliver exceptional member experiences.
