You are here: AI + innovation — what is implemented, what is concept, and where humans stay in the loop.
AI Lab
AI becomes valuable when it improves human capability
Not when it produces content faster. This is where I keep the implemented work, the prototypes, and the open questions — clearly distinguished from one another.
A learning pathway for agentic AI
I mapped where agentic AI can carry work, where a person has to stay in the loop, and what has to be true about governance and skill first. This is an operating architecture, not a deployed platform.
Examine the AI architectureStatus labeling
Nothing on this page implies deployment unless it is labeled as implemented.
AI + Learning
AI-assisted learning experiences where the technology improves practice, feedback, or access — not production speed alone.
- Implemented Work
AI-Enabled Training Toolkit
Task-anchored enablement resources and instructional approaches for workplace learners and trainers.
View - Concept / Prototype
Practice partner for conversational scenarios
Structured rehearsal with an AI counterpart, with human feedback retained as the assessment mechanism.
AI + Workflow
Performance and productivity workflows where AI removes genuine friction from work people already do.
- Concept / Prototype
Analysis-to-artifact workflow
AI-assisted drafting of job aids and reference from documented process, with human verification as a required step.
AI + Knowledge
Retrieval, resource systems, and assistants applied to organizational knowledge access.
- Concept / Prototype
Asynchronous assistance layer
Extension of the AOAS architecture: retrieval over owned, current organizational knowledge with explicit source attribution.
View
AI + Content
Responsible use of synthetic media, generative tools, voice, video, and rapid development.
AI + Strategy
Thought leadership on responsible organizational AI adoption.
Open space
What If? transformation studies
Publicly reported organizational problems analyzed as hypothetical engagements — diagnosis, approach, and what I would build. Clearly hypothetical, never presented as client work.
Contact
Working on AI adoption?
Most of what determines success is judgment, permission, and workflow fit — not the tool. I am glad to talk through where yours is stalling.