Should You Fill Your Next Vacancy With an AI Agent?

By Satish Thiagarajan, founder of Brysa, a Salesforce and data consultancy based in the UK. His company advises media, industrial, and services clients on using Data Cloud and Agentforce to turn signals into action. His work focuses on closing the loop between insight and execution in sales, marketing, and service.

When an employee leaves, most businesses would simply move to replace them. But, with AI agents increasingly capable of handling complex business processes, there is another option: reconsider the role itself.

Typically, jobs evolve over their lifetime. Responsibilities are added, processes evolve and technology changes the way work is completed. By the time an employee leaves a role, it can look very different from the one they were originally hired to fill.

As a result, vacancy can be a useful opportunity to reassess what a business actually needs. Which tasks still need to happen? Which could be automated? Where could an AI agent provide additional capacity? And which elements are always going to require a person?

The objective isn’t to simply replace an employee with an AI agent. It’s to make sure that you can meet the expectations, demands and obligations of the role as it stands today, not how it looked several years ago.

Start With the Work

Although resignation reduces headcount by one, that reduction doesn’t necessarily need like-for-like replacement by another full-time employee working in the same way.

Of the collection of responsibilities that constituted the role, some may require specialist experience. But some may be repetitive administrative tasks and others may no longer be necessary. Others still may be suitable for automation or an AI agent.

Breaking the role down like this opens up a range of possibilities.

Instead of asking whether an agent can replace an employee, businesses can identify individual areas of work where agents could provide additional capacity: customer enquiries, lead qualification, research, information processing or routine workflows. The result of analysis may still be a hire, but it could be a very different one.

Why AI Agents Change the Hiring Decision

Cost is an obvious consideration in the hiring decision, but it isn’t the only reason businesses are increasingly looking to agents.

Recruitment takes time and, even when successful, new hires need to be onboarded and provided a runway into the role. They’re not productive immediately and, as they bed in, they can place demands on other team members.

So far, so normal. But adding further human capacity as demand increases means repeating that process, and agents offer another way.

Once properly implemented, an agent can handle repetitive work as it arrives and scale more easily if required to. For the first time, increasing the amount of work doesn’t necessarily mean increasing the size of the team.

Enterprise Technology Is Making This Possible

New technologies are changing what can realistically be delegated to an agent.

Traditional automation is highly effective where a workflow follows predetermined steps. Agents go further: they can interpret information, make individual decisions within defined parameters and take actions across business systems to achieve their aims. Salesforce’s Agentforce is one example, enabling agents to work across enterprise workflows rather than simply generating information for a human employee to act on.

As these capabilities are integrated into mainstream business platforms, using agents within day-to-day operations becomes an obvious practical solution. The focus for businesses is therefore shifting from what the technology can do, to where it can be applied most effectively.

Design the Agent Around the Work

One mistake to avoid is to take an existing job description and adapt it as a specification for an AI agent. While human roles typically represent a mosaic of accumulated responsibilities, agent implementation provides allows you the opportunity to be much more precise.

Start with the outcome the agent is expected to deliver. Define which systems and information it needs to access, which decisions it can make independently, and which situations require escalation to a person.

Ensure access to applicable, high-quality information. An agent working from incomplete or fragmented customer records, outdated documentation or conflicting information will inherit those problems, and more data doesn’t mean better. It simply increases the risk.

Take ownership. A named individual within the business needs responsibility for reviewing performance, handling exceptions and changing the agent’s scope or permissions when required.

Success should then be measured against the outcome the agent was introduced to deliver: accuracy, resolution rates, escalation levels, cost per outcome or, perhaps, the amount of employee time released for other work.

When You Should Still Make a Traditional Hire

Some roles remain strongly dependent on human capabilities and involvement. A trusted account manager, for example, provides value through an established relationship as well as the tasks they complete. Leadership roles depend on accountability and the ability to make decisions when information is incomplete. Other positions rely heavily on experience, negotiation or an understanding of organisational context that can’t be reduced to a workflow, and in these cases, replacing a person with an agent risks removing the value the role was designed to bring.

A business may examine a vacancy, identify opportunities for automation and still conclude that another person is the right choice. Alternatively, it may find that agents can handle enough of the underlying workload that the role can be redesigned, reduced or removed altogether.

AI agents have therefore added another stage to the hiring process. Before asking who should fill a vacancy, businesses can first ask what actually needs replacing.