AI Staff Role Builder
Turn an unfilled business role into a trained, reusable AI agent
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 97%
The AI Staff Role Builder starts from an actual capability gap rather than from a fashionable tool. Map the roles a dream company would contain but the business has not hired, then select a bounded role that AI can plausibly assist with. Define the role's duties, output standards, company context, and domain knowledge, and use an AI tool to help draft persistent instructions. Review and store those instructions so the role can be called again without repeating the full brief. When a human team will use the agent, introduce it with a clear name, job description, and place in the workflow so people understand when to call it. Begin with limited tasks, check every output, and refine the instructions. The guest uses human-like names and profiles as an adoption tactic, not as proof the agent is a person.
Origin
Alicia Little explained on Aspire with Emma Grede how her company defines, trains, names, and introduces AI agents for specialist roles.
Core principles
- 01Start with a real role the business needs
- 02Define outputs and knowledge before deploying an agent
- 03Persistent instructions reduce repeated briefing
- 04Human adoption improves when the role is easy to understand
- 05AI should support higher-value human work
How to run it
- 1
Map missing roles
List the specialist roles the business needs but has not hired because of budget, availability, or timing.
- 2
Select suitable work
Choose a role with clear, reviewable outputs that an AI assistant can support.
- 3
Write the role instructions
Define what the agent does, how it should work, and what company and professional knowledge it may use.
- 4
Make the role persistent
Save the reviewed instructions in the tool's available memory or agent configuration so the role can be recalled later.
- 5
Onboard human users
Explain the agent's job, inputs, limits, and place in the workflow to the people expected to use it.
- 6
Test and refine
Run bounded assignments, review the outputs, and improve the instructions before expanding use.
In the wild
Alicia describes creating an AI team member named Penny. The tool receives instructions about Penny's sales-copy role, the company, the industry, and the expected writing style, then retains that role so Penny can be called for later copy assignments.
→ The recurring sales-copy brief is available without rebuilding all of the instructions each time.
A human team member calls on a configured press-release agent named Jerry. The agent asks five questions and produces a formatted draft based on the company's established press-release instructions.
→ The employee gets a structured first draft while the role remains understandable to the wider team.
Common mistakes
Starting with a novelty bot
A tool without a clear job and workflow may be ignored by the team.
Skipping human onboarding
People need to know what the agent does and how it makes their work easier.
Automating unsuitable decisions
The guest keeps relationships, compliance, and customer judgment anchored in human responsibility.
Is it for you?
Best for
Repeatable, reviewable knowledge tasks such as sales copy or press-release drafting.
Not ideal for
High-stakes decisions, relationship work, regulated actions, or tasks without human review and security controls.
From the transcript
“In my dream company, the roles that I'm not hiring for right now.”
From the episode
How to Monetize AI and Build the Life and Career You Want