The question raised in management meetings is usually: how many of this team do we no longer need? The sharper question is: whose job description needs rewriting? In most organisations that have been through this shift, headcount has not fallen; what changed is the mix of work each person does across a day.

The mechanism is straightforward. A job is a bundle of tasks, not a single indivisible unit. What moves to software is the rule-based and repetitive part: summarising, classifying, drafting, reconciling data. What remains, and rises in value, is the work that needs judgement, customer knowledge and accountability. The result is not the deletion of a role; it is a shift of weight inside it.

Which tasks move and which stay

• Rule-based work: The clearer the inputs and outputs and the higher the repetition, the more likely it is automated.
• Context-dependent work: Decisions that rest on knowing the customer, the case history and the interest of the organisation stay with people.
• Accountable work: Where a signature and responsibility are required, the output of the system is a recommendation, not a decision.

The skill that is genuinely scarce

In most organisations the scarce skill is not writing better prompts; it is judging the output. An employee who can tell where a generated draft is accurate and where it is confidently wrong creates the real added value. That skill does not come from a one-day workshop; it comes from repetition on real work with feedback. This is why training that stops at demonstrating the tool leaves no lasting effect.

Why training before tooling works

Organisations that hand the tool to a team before the training usually see two outcomes: a small group that races ahead by teaching itself, and a majority that quietly sets the tool aside. Reversing the order makes the difference. A short course in which each person practises on their own live work turns adoption from an individual choice into a team routine. The cost of that is trivial next to the cost of licences nobody uses.

Why it matters for Iranian SMEs

For an Iranian company facing constraints in hiring specialists, upskilling the existing team is the cheapest way to add capacity. An experienced employee who has worked with your customers and processes for years holds exactly the domain knowledge no model replaces; what is needed is a layer of tool skill on top of it. The reverse path, hiring someone fluent in the tools and teaching them the domain, is both more expensive and slower.

A 90-day upskilling map

• Days 1 to 30: List the weekly tasks of two or three high-volume roles and mark each one as rule-based or context-dependent.
• Days 30 to 60: For the rule-based tasks, run a short training session on the real cases of that team rather than generic examples.
• Days 60 to 90: Measure output per person against the pre-training figure and turn whatever worked into a written routine.

Recurring mistakes

• Not stating the goal; a team that does not know whether training is about adding capacity or cutting headcount will not cooperate.
• Generic tool training instead of practice on the live work of that specific role.
• Leaving the subject to one enthusiast and never writing a routine the rest of the team can follow.
• Measuring success by tool usage rather than by work output.

Three moves for this quarter

• List the weekly tasks of one high-volume role and split them into rule-based and context-dependent.
• Run one short training session on the real cases of that team.
• Measure output per person before and after the training.

Frequently asked questions

• Which skills hold their value?
  Judging output, knowing the customer and the domain, and being able to own a decision. These are not replaced by tools; they are amplified by them.
• Which role should we train first?
  The one carrying the most rule-based work; the effect turns into a number fastest there.
• What if the team resists?
  Resistance usually comes from ambiguity rather than from technology. Being explicit about the goal, and about which work is being taken off the team, does more than any workshop.

Takeaway

The right measure of this shift is not the number of roles removed; it is the ratio of time your team spends on rule-based work versus work that needs judgement. Organisations that move that ratio deliberately gain capacity without adding headcount, and keep the experienced people they already have.

Glossary

• Role redefinition: A change in the mix of tasks within a job, without removing the job.
• Rule-based work: Work with clear inputs and outputs and high repetition, which is simpler to automate.
• Domain knowledge: Accumulated understanding of customers, market and internal practice that is undocumented and lives in the experience of people.
• Upskilling: Adding new skills to existing staff rather than replacing them.
• Human in the loop: A pattern in which system output is approved by an accountable person before it takes effect.