Work is the evidence.
A lesson click earns no credit. The learner submits an artifact against a real source packet.
Enablement Command Center · Working application
From AI training to useful work.
Giving people AI access is a start. Knowing they can use it well takes evidence.
I built a connected workspace for coaching, assessing actual work, recording human review, and examining adoption. Every result has a path back to its inputs.
Independent portfolio application · Fictional company and sample records
CoachingAssessmentHuman reviewEvidence
A person makes the call.
The model assesses the work. An explicit review records approval or a return for revision.
Make the math inspectable.
Review time and rejected work count. Practice results stay separate from operating value.
The idea in under a minute
Help an organization put AI to work.
Read the walkthrough
Buying AI tools is one step. Helping an entire organization use them effectively takes a plan.
Bill Mabry built this working AI enablement framework to connect employee learning, everyday work, and business value.
Employees practice tasks relevant to their jobs—such as using AI to turn meeting notes into a clear management brief—with coaching and feedback along the way.
Leaders can see where teams need support, how AI adoption is progressing, and how potential time savings compare with the costs involved.
Those insights help them decide where to focus training and further investment.
This demonstration uses fictional company data to show an approach businesses can adapt to their own people and workflows.
A real saved run · Sep 12, 2026
Open the work behind the demo.
The organization and its operating records are fictional. This verification run used the provided flawed and strong sample drafts, real model calls, and an explicitly recorded demonstration review.
Submitted work
A decision brief with source support
Turn workshop notes into a brief that separates observations, proposed actions, missing information, and decisions still requiring approval.
Source packetVersioned artifactPolicy checks
Inspect the artifact and sources ↗Actual model assessment
100/100
AI rubric judgment
The model explains its findings with quoted evidence. A score alone grants no completion.
Read the assessment
The brief is well-structured and closely follows the source packet. It correctly distinguishes targets from results, preserves human decision rights, and flags unconfirmed items. One minor grounding note: the artifact cites [W6] in a way that references the restricted-field source without reproducing the restricted identifier, which is acceptable. The action table meets the P5 handoff requirements. Word count and format are within task scope. No blocking violations were found.
claude-sonnet-4-6 · Run receipt included
Recorded review decision
Approval has a rationale.
Checked revision 2 against W1–W6 and P1–P5: the 15% figure is labeled as a target, approval and staffing remain unconfirmed, and each next action has a role, timing condition, and verification step. The restricted identifier is omitted. Approved for this fictional practice exercise only; this is a recorded demonstration review, not an employer credential.
Sandbox reviewer: Demonstration reviewer. This is an entered demonstration role, not verified corporate approval.
My contribution
I built the application and the operating model.
I connected the learning experience, live coaching, assessment, review decisions, and evidence exports. I authored four source-based exercises and the rules that keep training activity separate from operating results.
Seven connected views
Overview, people and skills, practice, coaching, review, insights, and evidence.
A complete evidence trail
Versioned drafts, source quotes, model run receipts, explicit decisions, and portable reports.
Clear limits
An independent application with synthetic data. It demonstrates a working approach; it does not claim a client deployment, certified proficiency, or realized savings.
Let's talk about the work
Working through an AI adoption challenge?
The workflow, the people, and the evidence all matter.