SHAWN  MARTELL People · Organizations · Public service

AI systems, adoption & learning

Making AI useful in everyday work.

I design and operate AI-supported workflows, develop people's ability to use them, and help organizations decide where to invest.

I am a Prosci Certified Change Practitioner, having completed the certification program with an AI adoption focus.

See the work

Systems I design and operate

Connecting information with useful work.

Organizational knowledge

Turning project experience into shared insight

I created and run the Client Insight Framework and its recurring forum. The process connects conversations with project teams, AI-supported synthesis, human review, and action tracking so knowledge can inform decisions across an organization. Leadership has approved expanding it, with colleagues in growth and marketing committed to acting on what it surfaces.

Context carries forward between sessions, with operating instructions making responsibilities and follow-up explicit. People conduct the conversations, validate the interpretation, and decide what to act on.

Evidence and assessment

Making claims traceable

I built an evidence-assessment workflow that connects source records, structured information, retrieval, and a scoring rubric. It preserves where information came from and limits scores when supporting evidence is missing.

I rebuilt it with automated checks for extraction accuracy, rule enforcement, and regressions. I use it on real requests to find relevant evidence and make missing support visible, including answering a live request within thirty minutes. Source validation and human review remain part of the workflow.

Coaching and review

Giving feedback a clear standard

I built and used a coaching proof of concept that combines AI-supported feedback, self-review, and final human scoring against a rubric. Untested criteria remain unscored so people can see where more evidence is needed.

People and teams I develop

Building capability through learning and practice.

Completed learning programs

Staff fluency and deeper technical practice

My team authored and delivered an introductory AI learning series covering assistant selection, prompting, research and writing workflows, learning support, and privacy. We used commercially available tools so staff could apply the skills in everyday work.

I championed the program, developed and directed the team, protected time for delivery, and contributed readiness-learning design and facilitation.

I also led the AI Accelerator's program model, roadmap, and workshop design. Delivered with a technical partner, the completed bootcamp offered deeper, project-based practice. The two programs served different learning needs.

Shared practice

Making the work understandable

In the community of practice I lead, I demonstrated my AI approaches, then facilitated a structured conversation about how the team was using AI and what people were learning.

Using ADAPT's assessment phase, we then prioritized opportunities together. The team agreed on a set of shared AI assets, each with an owner, and a teammate delivered the first. Individual experimentation became shared, owned practice.

Decisions that support adoption

Choose the approach. Set the standards. Learn from use.

Solution design

Making guidance easier to use

For a federal client, I designed an AI-supported approach to learning and using a service design procedure. The design treats the authoritative procedure as the source and works through the client's approved AI tools.

The client adopted the approach and is expanding it through a SharePoint intelligence layer now in development, building on its existing adoption roadmap.

Evaluation and judgment

Use evidence to guide the next investment

I have used participant feedback and pilot assessments to evaluate learning programs and guide further investment. That means examining the learning need, the experience, the conditions for delivery, and what the evidence can actually support.

ADAPT framework

A practical framework for adoption.

I developed ADAPT to connect opportunity assessment, governance, workflow design, learning, and evaluation, and apply its components internally.

The five phases form a full adoption cycle or can be used individually, each with clear outputs and responsibilities.

Leadership support, communication through change, and equitable access to tools and learning run throughout.

  1. Assess

    Map demanding, recurring tasks and listen to staff. Prioritize opportunities by usefulness and feasibility.

  2. Define

    Set data boundaries, approve tools, establish quality standards, and agree on how to judge progress.

  3. Architect

    Document the workflow, including human review. Create prompts, templates, and a learning plan that help people use it.

  4. Pilot

    Test with teams doing real work. Compare results with the agreed measures and refine the approach around what happens.

  5. Transform

    Build effective practices into procedures, onboarding, and project routines. Keep a forum for sharing lessons and improving the work.

Working on AI adoption or team capability?

I'm interested in helping government and nonprofit teams connect useful systems, staff development, and sound judgment.

Get in touch