At Salesforce Connections in Chicago, we stepped onto the stage with an enormous declare: Agentic AI isn’t a future idea – it’s able to run your small business now.
Positive, it’s a straightforward sufficient declare to make, however we didn’t simply construct a chatbot or prototype and label it “AI”. We constructed a stay, end-to-end demo referred to as LEoptical, powered totally by Agentforce.
LEoptical is a totally orchestrated Salesforce stack, autonomously working:
- Advertising and marketing Cloud configured, developed and deployed by brokers
- Loyalty Administration as self-serve by clients by way of brokers
- Service Cloud + Information Cloud auditing, triaging, and shutting instances – agent-run
- Slack because the command middle, the place people monitor, work together with, and direct the brokers in actual time
And we didn’t simply demo this stack — we received first place the Agentic Advertising and marketing Showdown.
Watch the Agentic Advertising and marketing Showdown
Try ListEngage’s full prize-winning demo from Connections without cost anytime you’d like.
The explanation? It was sturdy, actual, and it pushed past “AI assistant,” into operational transformation. And it was actual.
I lead Utilized AI at ListEngage, a Salesforce accomplice centered on placing innovation into motion, constructing with clients, for purchasers. Once we checked out what Agentforce may do, we didn’t see a instrument. We noticed a brand new working mannequin.
This text isn’t a construct information. We’re not going to let you know create brokers step-by-step.What we’re going to do is present you what’s attainable – what an AI-powered enterprise seems like inside Salesforce, what we constructed to get there, and what questions try to be asking if you wish to head down the identical path.
The LEoptical demo we dropped at life was designed to really feel much less like a “proof of idea” and extra like a residing enterprise. It was constructed to indicate how Agentic AI can function throughout clouds, throughout groups, and throughout your entire buyer lifecycle.
Right here’s what we confirmed on stage:
Advertising and marketing Cloud: Brokers as full-stack entrepreneurs
We kicked issues off with Agentforce controlling Salesforce Advertising and marketing Cloud – not simply executing duties however managing the total lifecycle:
- Configuration: Spinning up new journeys, knowledge extensions and content material blocks
- Improvement: Writing and modifying HTML, AMPscript and SQL
- Deployment: Publishing belongings, launching automations and scheduling campaigns
The agent wasn’t some assistant clicking buttons. It was working as a platform admin, developer and marketer rolled into one, responding to natural-language prompts and orchestrating your entire system.
Loyalty administration: The AI-driven buyer expertise
Subsequent, we confirmed what occurs when Agentforce meets Loyalty Administration. Within the LEoptical demo, a buyer may:
- Enroll in a loyalty program
- Redeem factors
- Test profit eligibility
- Replace preferences
– all by way of an agent.
No help tickets. No portals. Simply on the spot, clever interplay, pushed by your buyer knowledge and surfaced by way of pure dialog. And most vital, the agent revered enterprise guidelines – no hallucinations, no rogue selections.
Service Cloud + Information Cloud: Brokers that deal with the heavy lifting
On the help facet, our brokers have been wired instantly into Service Cloud and Information Cloud.
They might:
- Audit and log instances with traceability
- Resolve Tier 1 points with pace and compliance
- Faucet into real-time unified profiles by way of Information Cloud to tailor actions
These weren’t static scripts – they have been dynamic, context-aware brokers educated on enterprise logic and guardrails.
Slack: The command middle
Lastly, we introduced all of it collectively in Slack, which we handled because the management tower for your entire Agentforce stack.
- Brokers reported in to its work group
- Customers may assign duties
- Notifications, audits, and actions have been piped in real-time
- And most vital: people stayed within the loop
This wasn’t AI going rogue. It was AI embedded in operations, with visibility, traceability and accountability in-built from the beginning.
What we confirmed wasn’t a set of AI options. It was an working mannequin — a world the place enterprise runs itself, with brokers as digital coworkers throughout each division.
That’s what received the judges over. And that’s what bought the viewers leaning ahead.
Find out how to clear up for AI sprawl
Many corporations face the problem of AI sprawl with out technique. The AI hype prepare is at full pace and each group is beneath stress to “do one thing with AI.”
In order that they do.
They …
… spin up a chatbot.
… construct a immediate for a marketer.
… plug in a big language mannequin (LLM) to assist with case summaries.
Positive, it’s thrilling at first, however then comes the screeching halt.
Why? As a result of with no technique, AI turns into sprawl: disconnected level options, no measurement, no orchestration and particularly no impression.
Right here’s what we’ve seen firsthand:
- Instruments with out roles: Brokers are created with no clear operate or accountability. Who governs them? What’s success?
- AI for AI’s sake: Options are constructed as a result of they are often, not as a result of they clear up an actual enterprise ache level.
- No connective tissue: Advertising and marketing does one factor. Service does one other. No person talks to Gross sales.
- Zero measurement: Most orgs can’t let you know if an agent saved them time, cash or buyer frustration.
The tip end result? AI turns into one other venture on the shelf as an alternative of a brand new manner of working.
We realized by way of LEoptical that the actual energy of agentic AI doesn’t come from constructing a sensible agentIt comes from treating brokers as a part of your working mannequin.
It’s not “The place can we add AI?”
It’s “The place can brokers take significant accountability for outcomes we already care about?”
The shift from function to operate, hype to behavior, is what separates profitable transformations from glorified demos.
Construct a strategic AI portfolio by “considering in bets”
Right here’s the uncomfortable fact: we don’t truly know which brokers will drive probably the most worth.
Each group is totally different. What works in Gross sales for one firm may utterly upend in Service at one other. A use case that saves time in a single area, may create chaos in one other.
So, how will we transfer ahead? We expect in bets.
This concept, impressed by Considering in Bets by Annie Duke, reframes the problem:
Don’t ask, “Will this agent work?”
Ask, “What’s the anticipated worth of working this experiment?”
Each agent is a small, calculated threat (wager) you absorb alternate for the potential of outsized worth. To wager sensible, nonetheless, you want knowledge. You want construction. You want the framework for evaluating, evaluating and prioritizing agent use instances.
That’s precisely what we’ve developed.
The enterprise worth framework
Total Influence Rating = i=1n(CSix Wi x TFi x TMi)+SE-TP
Break it down:
- Criterion Rating (1–5): How impactful is that this use case individually?
- Weight: How vital is that criterion to your small business immediately?
- Time Worth Issue: like cash, shorter-term worth is value extra
- Time To Worth Multiplier: Sooner wins get additional credit score.
- Synergy Impact: Does this agent amplify the worth of one other? That’s a bonus.
- Tradeoff Penalty: Will this agent create friction, duplication, or confusion? Deduct accordingly.
It’s Enterprise Evaluation, Developed
This isn’t new territory. Consultants have been doing enterprise evaluation for many years. What adjustments with agentic AI is the unit of research.
You’re not taking a look at methods. You’re taking a look at brokers — and treating every one as a residing, evolving a part of your operations.
You begin with discovery:
- The place are the friction factors?
- What components of the client expertise really feel damaged?
- Which inner processes sluggish us down?
Then you definately rating, prioritize and check.
Some bets will fail. That’s fantastic. They have been small.
Others will take off – and that’s the place you double down.
From agent sprawl to agent technique
Whenever you deal with brokers as a portfolio – not as a pile of prototypes – you unlock scale.
You acquire readability. You keep away from the AI graveyard of deserted demos.
And most vital, you create a repeatable decision-making framework,one which enterprise leaders can perceive, and one which AI leaders can belief.
Find out how to get began: Your first 90 days with Agentforce
We’re not going to let you know construct an agent. Not right here.
What we will do is let you know begin considering like a company that’s prepared for brokers to be greater than a proof of idea.
You don’t must boil the ocean.
You don’t should be a generative AI skilled.
You might want to know your small business,and be prepared to make sensible bets.
Right here’s what your first 90 days ought to concentrate on:
1. Begin with ache factors, not hype
What’s breaking inside your org immediately?
The place are clients ready, staff escalating, or groups saying “this shouldn’t be guide”?
If there’s no ache, there’s no urgency.
AI brokers succeed after they clear up actual issues.
2. Assume roles, not instruments
Don’t ask: “What can this tech do?”
Ask: “If I may rent a digital worker, what job would I give it?”
Each agent ought to have:
- A job title
- A accountability
- A KPI
- A supervisor (i.e., a human within the loop)
Whenever you outline roles this clearly, constructing turns into the straightforward half.
3. Orchestrate, don’t isolate
Your Advertising and marketing agent? It ought to discuss to your Loyalty agent.
Your Service agent? It ought to perceive holistic buyer profiles.
Your Slack command middle? It ought to have the ability to direct, monitor, and log all of it.
Brokers are teammates, not toys.
The second they function in silos, you’ve already misplaced.
4. Measure like a CFO, function Like a COO
You’re not simply launching a brand new function – you’re investing in a brand new mannequin.
Observe:
- Time to worth
- Operational impression
- Buyer expertise raise
- Synergy throughout processes
And use the enterprise worth framework from earlier to information prioritization.
5. Socialize early, iterate publicly
Don’t await perfection. Ship early, get suggestions, refine. When folks see brokers doing actual work — work they used to personal — they don’t panic. They lean in. They begin asking:
“May an agent assist me, too?”
That’s when adoption turns into exponential. Agentic transformation isn’t about automation. It’s about evolution.
And evolution begins with the way you assume, not what you construct.
That is just the start for AI brokers
The LEoptical demo wasn’t constructed to impress a panel. It was constructed to show some extent:
Agentic AI isn’t coming. It’s already right here –and it really works.
We confirmed what’s attainable when brokers aren’t bolted on however in-built.When Slack turns into a management middle.When Advertising and marketing configures itself.When Service Cloud instances resolve with zero-touch. When Loyalty is absolutely self-service.When the Salesforce platform runs—not as a result of of people, however alongside them.
That was the demo. However that is the second:
The second the place leaders cease asking “Ought to we use AI?”
And begin asking, “How may we use AI?”
As a result of the actual query isn’t if AI will remodel your small business. It’s whether or not you’ll lead that transformation — or scramble to catch up.
Let’s discuss for those who’re able to discover how Agentforce can operationalize AI inside your small business. We’ve seen what’s attainable. And we’re simply getting began.
Make your buyer interactions a two-way dialog
Say goodbye to “don’t reply” advertising. Learn the way Advertising and marketing Cloud Subsequent brings autonomous AI brokers, actionable knowledge, and full-funnel workflows collectively for really related engagement.
