Appian executive says AI adoption should start with the business problem, not the technology

Aldana said the wrong starting point is asking where AI can be used, because that puts the technology ahead of the business problem. Organisations should first work out what they are trying to solve and where the biggest bottlenecks and inefficiencies sit, he said, which means examining which decisions they want to improve, what data those decisions rely on and where human oversight needs to sit. That discovery work should involve the people who understand the process and its challenges, from IT and business teams through to executive leadership. "The test isn't whether AI can do something. It's whether it can make the process measurably better," he said.

On separating processes that genuinely need AI from those better served by conventional tools, Aldana said the key distinction is whether a task follows predictable rules and has a clearly defined outcome. Where it does, traditional automation or business rules can often deliver the result faster and more cheaply, and adding AI may not improve the outcome while introducing complexity that is not needed. He cited an insurance company that uses business rules to classify and route incoming online claims based on what a claimant selected from a drop-down list, sending automotive claims to one department and home insurance claims to another.

AI agents become more valuable, he said, when work requires adaptive reasoning, complex context or variable inputs. In the insurance example, an agent could analyse claims from web and mobile forms carrying structured data alongside incoming emails with unstructured content, and use keywords such as "motorway," "clash" and "passenger" to deduce that a claim concerns a vehicle accident and route it to the appropriate department. The mistake, Aldana said, is assuming that because an agent can perform a task it should. He argued for a balanced portfolio approach, using rules for high-volume, repeatable logic and agents for dynamic triage or investigation.

Asked about the hidden costs of applying AI to every process, Aldana said one of the highest lies in the operational infrastructure required around the model to make it work effectively and safely. Beyond the cost of the model itself, agents need access to the right data and systems, and businesses need monitoring, security and governance covering what those agents can access and what actions they can take. Accountability is a further question: organisations need to understand why an action was taken, trace what happened when something goes wrong and have a clear way of handling situations an agent cannot handle confidently. Without a unified platform to orchestrate data and enforce policy, he said, the overhead of managing exceptions and audit trails will outpace the productivity gains. That overhead can be worthwhile when AI delivers meaningful value, he added, but every additional agent introduces complexity, so businesses need confidence that the improvement to the process justifies it.

On where human judgement remains irreplaceable, Aldana said it stays critical where the cost or consequence of a wrong decision is high, particularly in regulated industries such as financial services, insurance, life sciences and the public sector. AI can accelerate areas including client onboarding, insurance, underwriting and clinical trials, he said, but there will still be decisions where the consequences mean a person remains involved.

Editor's Summary

Appian's Gregg Aldana told TechRadar that companies should pick AI use cases by starting from the business problem and measuring whether the process improves, rather than applying AI wherever it is technically possible. He said rules and conventional automation suit predictable, high-volume work, while agents fit tasks needing adaptive reasoning, and he pointed to monitoring, governance, security and exception handling as the main hidden costs of broad AI deployment.