The next phase of Salesforce AI is approaching quickly.
Agentforce is moving the conversation beyond AI assistance toward AI agents that can take action across service, sales and internal operations. For organizations already investing in Salesforce AI, the opportunity is significant.
But there is a growing problem.
The technology may be advancing faster than the skills organizations need to implement it.
As businesses move from AI experimentation toward real-world deployment, they need people who can do far more than configure new Salesforce functionality. They need teams capable of designing agent workflows, connecting data, integrating systems, managing security, establishing governance and measuring whether AI is actually delivering value.
That makes the emerging AI skills gap one of the biggest challenges facing Salesforce leaders.
The data suggests that pressure is already building. According to the Mason Frank Careers and Hiring Guide 2026, 59% of hiring managers have seen demand for AI specialists increase over the past year, while 48% say AI-related roles are harder to fill than other technology positions. At the same time, 66% of Salesforce professionals plan to pursue AI training or certification in the coming year.
That combination tells an important story. Organizations are trying to build AI capability at the same time as the Salesforce workforce itself is racing to develop it. For leaders moving Agentforce from experimentation into production, waiting for the market to catch up creates a real delivery risk.
And with Dreamforce approaching, the scale of that challenge is becoming increasingly visible.
The Agentic Enterprise will be a major theme at Dreamforce this year, but the agenda goes far beyond product innovation. Sessions covering AI-ready data, governance, architecture, human adoption, AgentOps and the changing responsibilities of Salesforce professionals suggest that building the workforce behind AI will be just as important as deploying the technology itself.
Building an agent is only the beginning
Agentforce makes it possible to introduce AI agents into a growing range of Salesforce workflows.
Customer service teams can use agents to handle routine requests. Sales organizations can automate repetitive activities. Internal teams can use AI to retrieve information, complete tasks and support employees.
The technology can create significant efficiencies.
But moving from a working agent to a reliable business process requires a much broader set of capabilities:
- Someone needs to understand the workflow being changed
- Someone needs to determine when human intervention is necessary
- Someone needs to decide which actions the agent should be allowed to take
- Someone needs to connect the systems and information required to complete those actions
- Someone needs to measure whether the new process is actually performing better than the one it replaced
This is why Agentforce adoption cannot be treated simply as another Salesforce implementation project.
Organizations need a combination of technical and operational expertise capable of turning agent functionality into processes that work reliably within the business.
Dreamforce will explore this challenge directly. Sessions such as ‘From AI Pilots to Enterprise Impact’ and ‘Turning AI Pilot Sprawl into Measurable Agentic Value’ will examine how organizations move beyond experimentation and create the governance and delivery models required to scale working agents across the enterprise.
AI-ready data becomes a workforce challenge too
AI agents depend on context.
If an agent is expected to support a customer, recommend an action or complete a business process, it needs access to reliable information.
That makes data one of the most important foundations of the Agentic Enterprise.
Many organizations already have large volumes of customer data. The challenge is whether that information is connected, accurate and usable across the workflows AI needs to support.
Different teams may maintain different customer records. Important information may sit outside Salesforce. Data ownership may be unclear. Permissions may prevent systems from accessing the information they require.
AI exposes those weaknesses quickly.
Solving them requires more than Data Cloud knowledge alone.
Organizations need people who understand data architecture, integration, governance and business context. They need professionals capable of determining which information an agent needs, where that information should come from and how it should be maintained.
This makes AI readiness a multidisciplinary challenge.
Dreamforce's agenda reflects that reality. ‘Build Your AI-Ready Data Strategy’, ‘Data Leaders Roundtable: Trust, Strategy, and AI’ and ‘Agentic Data Engineering: Automating Trusted Data for AI’ will all examine the data foundations required for production AI, including integration, quality, governance and trust.
Mason Frank works closely with organizations across the Salesforce ecosystem to understand where these capability gaps are emerging, particularly across architecture, integration and data. This insight can help leaders identify where existing teams need support and where additional expertise may be required as Agentforce moves into production.
Governance needs to be designed into AI from the start
AI agents introduce a fundamentally different governance challenge.
Traditional automation follows predefined rules. Agents can operate with greater autonomy, interpret information and take actions based on context.
That makes oversight critical.
Organizations need clear decisions around what agents can access, which actions they can complete and where human approval remains necessary.
They also need processes for monitoring performance and identifying unexpected behavior.
These responsibilities cannot sit solely with security or compliance teams.
Salesforce architects, Admins, Developers, Consultants and business stakeholders all need to understand how governance affects the systems and workflows they manage.
This creates demand for professionals who can balance innovation with control.
The goal is not to restrict AI so heavily that it cannot create value. It is to establish guardrails that allow organizations to expand AI confidently.
That balance will be another major Dreamforce theme. ‘Making AI Work: Governance and Guardrails That Matter’ will explore governance models, ethical guardrails and testing approaches for Agentforce, while ‘Who's Accountable When AI Gets it Wrong?’ will examine the trust and accountability foundations organizations need to scale AI safely.
Existing Salesforce roles will need to evolve
The AI skills emergency does not mean every organization needs to create entirely new roles.
In many cases, existing Salesforce positions will evolve:
- Admins may increasingly manage automated and agent-driven workflows alongside traditional platform administration
- Developers may spend more time building integrations and scalable AI-enabled solutions
- Architects will need to consider how agents interact with data, applications, security models and enterprise architecture
- Consultants will need to help organizations redesign processes around AI rather than simply implement functionality
- Business Analysts may become increasingly important in identifying where agents can create value and translating business requirements into AI-enabled workflows
The strongest Salesforce teams will therefore combine existing platform expertise with new AI capabilities.
Mason Frank helps organizations understand how these changing capability requirements translate into their Salesforce teams, providing access to specialist expertise across AI, data, architecture and transformation.
Organizations that invest in developing these skills internally may be able to create significant capability without rebuilding entire teams.
But that requires leaders to identify where roles are changing before capability gaps begin affecting delivery.
Dreamforce is putting this evolution directly on the agenda. ‘Your Evolved Role: Human Skills for Admins Managing Agents’ explores how Admins will delegate work to agents and manage virtual teammates, while ‘A Day in the Life of an AI-Assisted Architect’ and ‘Architect Salesforce in the AI Era’ examine how architecture, security, DevOps and decision-making change when AI becomes part of everyday Salesforce delivery.
Technical AI capability alone will not be enough
Some of the most important skills required for the Agentic Enterprise will not be purely technical.
Organizations also need people who understand processes, customers and commercial objectives.
An AI agent may be technically capable of completing a task, but that does not automatically make the task worth automating.
Teams need to understand where automation improves customer experience, where it creates operational efficiency and where human judgment remains more valuable.
That requires collaboration between Salesforce professionals and the business.
It also requires people who can challenge assumptions:
- How will success be measured?
- Should this process be automated?
- How will employees work alongside it?
- What happens when the agent cannot complete the task?
- Does the agent have enough information to make this decision?
These questions sit at the intersection of technology and business transformation.
The professionals who can operate effectively across both will become increasingly valuable as organizations expand their use of AI.
Dreamforce will reinforce this human dimension too. ‘Powered by You: The Human Layer of the Agentic Enterprise’ focuses on the tools, training and confidence people need to embrace AI, while ‘TIME100 AI: The Human Side of Scaling AI’ will explore the workflows, guardrails and judgment required to build trust as AI moves into production.
AI skills need to become an ongoing capability
Salesforce AI will not stand still.
Agentforce will evolve. New agent capabilities will emerge. Data requirements will change. Governance expectations will mature.
That means organizations cannot treat AI skills as a one-off training requirement.
They need a model for continuous capability development.
Teams will need opportunities to experiment with new functionality, understand changing best practices and apply what they learn to real business challenges.
Organizations will also need to think differently about workforce planning.
Rather than asking only which roles they need today, Salesforce leaders should consider which capabilities they will need over the next two or three years.
That may include:
- AI governance
- Agent optimization
- Data architecture
- Integration expertise
- AI workflow design
- Business process transformation
The objective is not to predict every future Salesforce role.
It is to build a workforce capable of adapting as the platform changes.
That theme is particularly visible in the Dreamforce session ‘5 Roles Every Enterprise Needs to Scale AI Successfully’, which focuses specifically on the capabilities required to drive adoption, scale AI responsibly and ensure AI strategy creates value across the business.
The biggest Agentforce challenge may be people, not technology
The Agentic Enterprise creates a significant opportunity for Salesforce organizations.
AI agents could change how customer service operates, how employees complete routine tasks and how businesses use the information already held across their Salesforce environments.
But technology alone will not deliver that transformation.
Organizations need people capable of turning AI functionality into reliable business processes.
They need architects who understand AI-ready environments, data specialists who can create trusted foundations, professionals capable of governing agents responsibly, and business-facing Salesforce experts who understand where AI can genuinely improve outcomes.
Dreamforce will give Salesforce leaders a clearer picture of what the Agentic Enterprise could look like. The more important question may be whether their organizations have the capabilities required to build it.
The AI skills emergency has already started.
Organizations that identify their capability gaps now will be better positioned to move from experimentation to production as Agentforce evolves.
Those that wait may discover that access to the technology was never the biggest barrier.
Their ability to implement it was.
Is your Salesforce team ready for the Agentic Enterprise?
Agentforce is changing the capabilities organizations need across architecture, data, governance and AI delivery.
Download the Mason Frank Careers & Hiring Guide to understand how Salesforce skills are evolving and identify the expertise your organization may need as AI adoption accelerates.