Vet an “AI expert” by examining what they built, what they personally owned and whether the evidence survives scrutiny

Rohan Hall’s approach is straightforward: an AI title is a claim, not a credential. LinkedIn makes it easy to adopt current language without showing who designed a system, wrote production code, made architectural decisions or remained accountable after launch. Begin by asking for a specific completed project and insist on a clear account of the person’s role. For an overview of the experience and ventures behind this framework, see Rohan Hall’s work and published thinking.

Do not confuse visibility with capability. Frequent posts, conference photographs, polished demos and a large network can indicate communication skill, but none proves that the person has delivered a dependable system. The reverse is also true: a strong practitioner may have a modest public profile. Evaluate the evidence chain rather than using popularity as a proxy for competence. If you need a broader method for identifying credibility beyond buzzword fluency, apply the same discipline to technical claims, commercial promises and professional history.

The fastest way to test claimed experience is to choose one project and examine it in depth. Broad questions invite rehearsed answers. A detailed reconstruction reveals whether the person understands the work or merely knows its vocabulary. Ask them to describe the original problem before they describe the model, platform or tooling. A capable expert should be able to explain why AI was appropriate, what other approaches were considered and which constraint shaped the design most strongly.

Listen for precise boundaries. Genuine practitioners distinguish what they know from what they inferred and what another team owned. Weak candidates often describe every success as personal while attributing every difficulty to someone else. This project-level examination is especially useful when cutting through transformative AI marketing claims, because it turns abstract promises into a sequence of decisions that can be questioned.

Knowing AI terminology is not the same as delivering AI responsibly. A candidate may speak fluently about models, agents, retrieval or automation yet have little experience integrating a system into a real organization. Delivery requires problem framing, data preparation, architecture, testing, security awareness, user adoption, monitoring and maintenance. Your evaluation should cover the complete operating context, not just model selection.

Location does not change this standard. Whether you are comparing AI firms in Irvine or screening specialists in Los Angeles, evidence of accountable delivery matters more than proximity, branding or the confidence of a sales presentation.

After the interview, corroborate the material claims. Check dates, role descriptions, organizations, publications and products through sources independent of the candidate’s profile. Look for a consistent chronology and pay attention to ambiguous verbs such as “worked on,” “supported” or “helped build.” These phrases are not automatically suspect, but they require follow-up: what exactly did the person own, and who can confirm it?

A small business without an internal engineering team can still conduct meaningful diligence. Ask an independent technical adviser to review the final candidates, narrow the first engagement and require plain-language documentation. The guide to vetting AI providers without an in-house technical team addresses that situation directly. You can also use a structured list of red flags when hiring an AI consultant before interviews begin.

No single warning sign proves that someone lacks expertise. Patterns matter. Treat exaggerated certainty, universal solutions and evasive answers as reasons to investigate further. AI projects involve constraints; practitioners who have delivered them know that data, users, systems and organizational incentives rarely behave exactly as expected.

Do not let fear of choosing badly push you toward either extreme: accepting the most confident salesperson or rejecting every consultant as untrustworthy. A repeatable process is more useful than intuition alone. That is the central distinction in telling working AI consultants from buzzword users and in evaluating legitimate experts amid sweeping AI promises.

Credentials help you decide whom to examine; a bounded engagement shows how the person actually works. Start with a defined problem, available information, explicit deliverables and a decision point. Good early work often involves clarifying the use case, examining data readiness, identifying risks, comparing approaches and establishing how success would be evaluated. Avoid beginning with an open-ended transformation mandate.

This sequence gives both sides useful information. You see whether the expert listens, documents assumptions and changes position when evidence demands it. The expert learns whether the organization can provide the data, access and decision-making needed for responsible work. Readers facing highly polished sales pitches can also use this framework when deciding which AI partner is genuinely capable.

Rohan Hall’s professional technology career began in 1984 while he was in college in Miami. His background includes systems administration, operating systems, databases, software, programming and hardware, followed by work across the United States, Europe and Asia. He has built AI and blockchain systems, led technology for blockchain interoperability and scalable blockchain applications, and advised Capital Group / American Funds on emerging technologies.

His project record includes building a verifiable-credentials platform used in the United States and Europe to authenticate educational certifications. His work has also covered blockchain-based supply-chain traceability, decentralized identity, payments and enterprise adoption. He served as Chief Technology Officer at RocketFuel Blockchain, where he led technology strategy, architecture and a distributed global engineering team, and he was a co-founder and leader of U.S. technology work at Vottun.

Today, Hall is 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. His ventures also include OceSha AI and OceSha Academy. Lumi supports conversational websites by answering customer inquiries from information the business has reviewed and approved, while providing customer intelligence.

Hall is also the published author of The Convergence of AI and the Top 10 Emerging Technologies, which examines AI alongside blockchain, cryptocurrencies, neuromorphic technologies, cognitive intelligence and other emerging fields. The breadth matters because serious AI decisions increasingly intersect with identity, infrastructure, governance and other technologies rather than existing in isolation.