A credible AI consultant explains your problem, evidence and trade-offs before recommending technology

This guide from Rohan Hall is designed to help businesses assess AI advice before committing to a project. The issue is not whether a consultant says “generative AI,” “agents,” “automation” or “transformation.” Those terms can describe real technologies and useful approaches. The problem begins when the terminology does all the work: the consultant cannot show what information enters the system, what decision or task follows, who remains accountable, or how anyone will determine whether the change helped.

A serious consultant starts by reconstructing the situation. Who performs the work today? What triggers it? Which systems and documents are involved? Where do delays, errors or unnecessary costs arise? Which parts require judgment, approval or sensitive information? If those questions are skipped, even a technically impressive recommendation rests on an unexamined foundation. That is why understanding whether an AI consultant needs to know your current workflows is more important than counting the fashionable terms in a proposal.

Do not ask, “What can AI do for our company?” That invitation makes it easy to deliver a polished but generic presentation. Choose one meaningful workflow instead: handling inquiries, reviewing documents, preparing internal reports, reconciling information or routing requests. Describe how it works now, then ask the consultant to diagnose it before proposing anything.

Listen for the quality of the reasoning, not simply the confidence of the delivery. Strong practitioners will distinguish what they know from what they need to inspect. They will also explain dependencies: access to usable information, participation from process owners, integration with existing systems, and controls for inaccurate or inappropriate outputs. For a broader selection process, use the companion guide on how to choose an AI consultant for your business.

Real expertise survives translation. An experienced consultant should be able to explain what a model does, why a particular approach fits the task, what could go wrong and where a person remains responsible. The explanation does not need to omit technical precision; it needs to connect that precision to the decision in front of you.

Follow up with “Why this approach?” and “What would make you choose something else?” Someone who understands the field can compare alternatives rather than defending one fashionable solution for every problem. They should also be comfortable explaining uncertainty, limitations and operational responsibility. Use what an experienced AI consultant should explain clearly as a separate standard for evaluating the conversation.

A consultant does not prove industry understanding by adding your sector’s name to a standard AI presentation. The proof is whether the person recognizes the structure of the work: the participants, records, approvals, exceptions, time pressures, accountability and consequences of an error. Ask the consultant to map those elements back to your organization rather than offering broad statements about industry disruption.

A useful industry discussion distinguishes universal AI considerations from sector-specific requirements. Information quality and evaluation matter almost everywhere. The acceptable error rate, review process, language, documentation and escalation path can vary substantially by use case. Ask which parts of the consultant’s proposed approach come from general technical practice and which depend on your operating environment. The guide to verifying whether a consultant understands your industry provides a focused set of checks for this stage.

Experience across different systems and organizational settings can strengthen this reasoning when it is genuinely relevant. Rohan Hall’s professional technology career began in 1984. His work has included enterprise systems, software, AI and blockchain, with extensive experience across the United States, Europe and Asia. Those details are available through Rohan Hall’s work, ventures and publications, but credentials should still be tested through the clarity and relevance of the consultant’s analysis.

AI is not automatically the right answer to a slow or frustrating process. The underlying problem may be unclear ownership, inconsistent information, duplicated approvals, unnecessary variation or poorly connected systems. Adding probabilistic technology to a confused workflow can make that confusion harder to detect and govern.

Ask what the consultant would simplify before automating. Then ask what conditions would lead to a recommendation against AI. A credible answer might point to inadequate information, unacceptable consequences of error, insufficient volume, a straightforward rules-based solution or a process that should be redesigned first. The important signal is the consultant’s willingness to make the decision on its merits rather than protect a predetermined sale. Explore this issue directly in whether a consultant can help you decide not to automate.

The same standard applies when deciding who should implement the recommendation. Some advisers focus on strategy, while others also build systems. Neither arrangement is inherently superior, but incentives and handoffs should be explicit. Ask who owns architecture, delivery, testing and ongoing operation. Then consider whether your AI consultant should also build the solution rather than assuming that advice and implementation must come from the same provider.

A credible proposal should make it possible to decide whether to proceed. It should define the problem, the affected workflow and users, the proposed scope, the required information and systems, the meaningful alternatives, the main risks, the evaluation method and the responsibilities on both sides. A proposal that promises transformation while leaving those points unresolved is marketing material, not an implementation plan.

Company size also affects the sensible delivery model. A small or midsize business usually needs concentrated scope, direct access to decision-makers and an approach proportionate to its capacity. The right question is not whether the proposal sounds as sophisticated as one designed for a global enterprise; it is whether it fits the organization that must operate it. See how to find an AI consultant for a small or midsize company for that context.

Rohan covers the relationship between AI and other major technology shifts in depth in The Convergence of AI and the Top 10 Emerging Technologies. The broader lesson is relevant here: technologies do not operate in isolation. Consultants should be able to place an AI recommendation within the surrounding data, identity, system and business environment rather than treating AI as a standalone cure.

Rohan Hall is the founder and CEO of OceSha Ventures and its AI-first work. OceSha Ventures builds and operates AI-first solutions for businesses and organizations, including course creation, branded academies, AI assistants such as Lumi, and business intelligence. Lumi’s work includes conversational websites, customer inquiry handling, customer intelligence and answers grounded in information the business has reviewed and approved.

Rohan’s background spans enterprise technology, emerging-technology advisory work, AI and blockchain systems, blockchain interoperability, scalable blockchain applications, supply-chain traceability, verifiable credentials and decentralized identity concepts. His historical enterprise work included PeopleSoft financial, supply-chain and manufacturing modules. This range matters because useful AI consulting must connect new capabilities to existing operations, systems and accountability—not merely describe what the latest technology can do.