AI iNITIATIVES | discovery, Analysis, and program leadership
AI creates value when it fits the way the organization actually works.
Successful AI initiatives connect a real business need with the data, workflows, technology, governance, and people required to support it. I help organizations define those connections, align business and technology teams, and move from opportunity through requirements, solution development, adoption, and delivery.
What effective AI initiatives need
A clearly defined problem.
Start with the decision, workflow, customer or employee need, and measurable outcome rather than with the desire to “use AI.”
Clear ownership and accountability.
Define who owns the outcome, who can make decisions about the solution, and who remains accountable when AI contributes to a decision.
Fit-for-purpose data.
Understand what data exists, what is trustworthy, what is missing, how it needs to be structured or integrated, and whether it can support the intended use.
Governance by design.
Bring risk, privacy, compliance, accessibility, security, and appropriate human oversight into solution design rather than adding them after the fact.
Workflow and adoption design.
Define how work will actually change, where AI enters the process, what remains human, and what people need in order to adopt the new way of working.
Business and technology iteration.
Evaluate real outputs together, refine requirements and solution behavior, and continuously improve accuracy, usefulness, efficiency, and fit with the underlying workflow.
A short discovery, pointed at an AI decision.
Establish what is actually true today: which decisions and workflows are candidates, what the supporting data can and cannot sustain, where ownership and governance gaps sit, and which regulatory, accreditation, privacy, security, and accessibility constraints shape what is permissible. The result is a shared current-state picture from which an investment decision can be made — not a justification for a capability someone has already decided to buy.
Before an investment
Establish whether the business problem, workflow, data, ownership, governance, human accountability, and constraints support proceeding before a larger commitment is made.
After an initiative stalls
Establish what prevented a promising capability from producing operational value, what can be corrected, and whether the organization should continue, redirect, test differently, or stop.
What the discovery produces
A clearly framed business decision, workflow, or need
A current-state workflow and human decision picture
An assessment of relevant data, gaps, and dependencies
Ownership, accountability, and governance questions requiring resolution
A proposed boundary between automation, human review, and human judgment
Decision criteria and a recommended next step: proceed, test, redirect, or stop
Typically two to four weeks, depending on the number of workflows, systems, and stakeholder groups involved.
How I help
AI opportunity and requirements definition — Translate a business problem into use cases, requirements, decision logic, data needs, constraints, and measurable outcomes.
Workflow redesign and iteration around AI — Map how work happens today, determine where AI adds value and where human judgment must remain, define the target-state workflow, and bring business and technology teams together to evaluate real outputs and refine requirements and behavior against it.
Data and integration definition — Working with data and engineering specialists, define the data, integrations, and platform capabilities an AI solution depends on, so the investment is scoped against what the data can actually sustain.
AI delivery and program leadership — Coordinate the people, decisions, dependencies, vendors, data, and technology needed to move an AI initiative from discovery through implementation, with the analysis carried inside the leadership.
The goal is not to automate more. It is to improve how work gets done.
AI can accelerate synthesis, pattern recognition, drafting, analysis, and routine decision support. The more important question is where those capabilities belong in the work. Some activities can be automated, some require review, and others depend on context, judgment, trust, or accountability that should remain human.
Designing that boundary is part of designing the solution itself. It requires understanding the work, the people doing it, the decisions being made, and the consequences when the system is wrong.
AI drafts. Humans understand.
My role in the work
I lead the business-side discovery, analysis, alignment, requirements, and program structure around the decision. I work with data, engineering, security, privacy, legal, accessibility, and subject-matter specialists where their authority and technical depth are required. The work is to connect those disciplines into a decision and delivery path, not to substitute for them.
How AI fits into my own practice
I use AI to accelerate synthesis, pattern recognition, knowledge organization, modeling, workshop preparation, and first drafts of analysis and communication. That creates more time for the work that requires human judgment: asking the right questions, challenging assumptions, facilitating alignment, interpreting context, and making decisions.
Working this way also provides practical insight into where AI genuinely improves knowledge work and where human involvement remains essential.
Considering an AI investment, or recovering from one?
A short discovery can establish where the business need, workflow, data, ownership, governance, and adoption conditions actually stand, before more budget is committed, or after a pilot that worked in the demo and changed nothing downstream. The first decision may be whether the initiative should proceed at all.