Work Samples

Proof of how I think through practical AI problems.

These examples focus on turning unclear work into structured information, identifying realistic opportunities, and building tools that support better decisions.

AI Process Map Builder

A working Firebase-hosted tool for collecting ordered workflow steps, ownership, tools, effort, friction, and risk signals. It produces a readable submission for review while keeping the visitor's mental model simple.

Open the Process Mapper

AI agent pilot proposal

Context: large health insurer. Outcome: proposal approved for a pilot estimated at up to 40,000 labor hours saved annually. Evidence: private work sample; the proposal and approval context are available for discussion.

FAQ and knowledge helpers

Artifact: source-grounded assistant work. What it shows: turning approved reference material into a usable question-and-answer experience. Evidence: documented in the James Lane Assistant repository.

Interview and workflow tools

Artifact: structured interview and workflow exploration tools. What it shows: using small web interfaces to turn vague inputs into clearer next steps. Evidence: related implementations and documentation are collected in the James Lane Assistant repository.

See the reasoning in action

Start with a workflow you already understand.

The mapper does not pretend to replace domain expertise. It creates a clearer starting point for asking better questions.

Build a Process Map