AI strategies are often developed at leadership level before the organisation has examined how work is actually undertaken.
Senior direction remains essential. Leadership must establish priorities, allocate resources and define the organisation's appetite for change.
However, many of the most valuable opportunities are identified by the people dealing with inefficient systems, repetitive tasks and information gaps every day.
Staff understand where information is repeatedly re-entered, where approvals become delayed, where project knowledge is difficult to locate and where existing systems fail to reflect the realities of the work.
A bottom-up approach begins by engaging teams across the organisation.
This should not be limited to asking staff which AI tools they would like to use. The more useful discussion concerns their work:
- Which tasks consume disproportionate amounts of time?
- Where are errors or inconsistencies most likely to occur?
- What information is difficult to access?
- Which activities rely heavily on individual knowledge?
- Where do systems create additional work?
- Which stages require professional judgement?
These discussions help separate genuine organisational needs from assumptions about what AI might do.
The objective is not to automate everything. Some activities should remain deliberately human because they depend on judgement, accountability, negotiation or professional relationships.
The purpose is to understand where technology can remove avoidable friction and allow people to concentrate on higher-value work.
Once a suitable opportunity has been identified, staff should remain involved throughout design and testing.
Early prototypes can be introduced into a controlled workflow, enabling users to test outputs, identify limitations and shape the solution before wider implementation.
This approach produces better systems because the operational knowledge of the organisation becomes part of the design.
It also supports stronger adoption. Staff are more likely to engage with a solution they have helped define than one introduced without reference to their working practices.
Bottom-up innovation does not mean an absence of strategic control. It means combining leadership direction with practical knowledge from across the organisation.
The strongest AI solutions are rarely introduced fully formed. They are developed through structured collaboration with the people who understand the problem best.