Artificial intelligence is often associated with large-scale transformation programs, sophisticated digital infrastructure and significant investment.
But for mid-sized businesses, some of the most valuable opportunities may be much simpler: solving clearly defined problems using the information, systems and technology they already have – and capturing additional data where it can add value.
Vectr Labs director Eben van Niekerk said businesses considering AI should start with the operational challenge rather than the technology itself.
“We start by finding out where the biggest bottlenecks and pain points are,” he said.
“Then we look at what data is already available, what might be missing, and create the simplest solution that can deliver the outcome.”
That approach shifts the focus from adopting AI for its own sake towards identifying where it can deliver a measurable business outcome.
Across sectors, businesses already generate large volumes of information through operational systems, spreadsheets, reports and other software platforms.
But Mr van Niekerk said businesses should not assume AI is limited to working with information already captured digitally.
“In many businesses, important operational information still sits in manual forms, documents, photographs, inspections or other unstructured sources,” he said.
“Increasingly, this information can be captured relatively easily and cost-effectively using tools such as digitised workflows, AI-based document and image processing, and computer vision.”
The approach is to understand what data already exists, identify what is missing and capture additional information where it adds value before testing the solution and scaling it.
For mid-sized companies, this creates an opportunity to introduce AI without replicating the technology environments of much larger organisations, potentially leapfrogging parts of the traditional digital transformation journey.
While the approach is particularly relevant for mid-sized businesses, the same principle can apply to larger organisations looking to move quickly on targeted use cases.
Mr van Niekerk said advances in AI had reduced the time and cost involved in developing digital solutions.
In mining, Vectr Labs has worked with mid-tier operators in complex environments where fragmented systems, operational data and technical constraints can present challenges.
“A mid-tier miner doesn’t need to recreate a tier-one company’s entire digital infrastructure,” he said.
“They can potentially leapfrog some of that infrastructure and move more quickly.”
Mr van Niekerk said mining provided practical evidence of an approach that could apply more broadly across industries.
Businesses can identify a specific bottleneck, understand the information available, capture what is missing where necessary and test whether AI can improve the outcome before committing to broader investment.
One area with significant potential is manual work that absorbs the time of highly skilled employees.
In mining and engineering, for example, employees can spend considerable time transferring information between spreadsheets, searching through studies, reconciling documents and assembling reports.
“The point is not that AI should write a feasibility study,” Mr van Niekerk said.
“It’s about allowing engineers to spend more of their time engineering, rather than carrying out manual tasks.”
The potential gains can be significant. In one recent example, a company using Vectr Labs’ Synapse reporting platform reduced the time taken to produce its quarterly board reporting from around two weeks to approximately one day.
Mr van Niekerk said the example demonstrated that the value of AI was not necessarily about replacing people but removing manual work and allowing skilled employees to focus on higher-value activities.
The task may differ across industries, but the underlying principle is similar: use technology to support employees by reducing lower-value manual processes and making information easier to capture, access and use.
That approach can also change the way companies think about the return on AI investment.
Instead of committing significant capital to a broad transformation before knowing what value it will create, businesses can begin with a defined use case, understand the data required, test whether the solution improves performance and expand from there.
As AI becomes embedded in more business processes, questions around oversight and accountability remain important.
“AI will make mistakes, but people also make mistakes,” Mr van Niekerk said.
He said businesses already relied on people to make decisions using information that could be incomplete or outdated, and the level of oversight should depend on the risks involved.
“Human accountability remains essential,” Mr van Niekerk said.
“AI should help people make better decisions, rather than replace the need for human judgment.”
For mid-sized companies, the opportunity may therefore be less about pursuing the most sophisticated AI capability and more about identifying where the technology can solve a genuine business problem.
Rather than waiting until they have the budgets, systems or infrastructure of much larger organisations, businesses can start with a defined problem, test the value of a solution and build from there.
As AI tools become more accessible, that could allow mid-sized businesses to leapfrog parts of the traditional digital transformation journey and focus investment where it can make the greatest difference.
