Phase One: Mapping Your World

This is Unit Four of The M-CaReM-AI Method, on phase one: mapping your world. Unit Three set out five phases. This unit does the first in depth: naming what a world is made of, before anyone describes what to do to it. Carry one question through this unit: how do you find the things nobody mentions because nobody touches them every day? The answer arrives with four interview questions. By the end you will be able to say what phase one is for and where things come from. To build an Object Inventory with zones. To avoid the common mapping pitfalls. And to run interviews and a workshop and draw the map.

Six entries on what phase one aims for, and where, and from whom, the things come.

A question nobody had asked.

MCM 076 The Auditor's Question

In the book's example, a benefits office lists what its new system must handle, until a compliance officer asks what it will show a judge in two years. Two notes go up: Policy Version and Case Snapshot. The things nobody touches daily are often the ones you most need.

What phase one aims for.

MCM 077 What Phase One Is For

Phase one aims for a good-enough shared picture: ten to thirty things, core separated from supporting, a simple map, and stakeholders saying that is basically our world. The book's image: sketch the coastline before the streets.

What it does not try to do.

MCM 078 What Phase One Is Not

Phase one does not define every field, settle every nuance, or lock anything down. It gives a language of shared nouns, a rough shape, and a starting point. Aiming for precision here stalls the work it is meant to start.

Where things are found.

MCM 079 Sources to Mine

The method names four sources: policy and regulation, existing system material, operational documents, and existing requirements. Paper forms and shadow spreadsheets often reveal things the database never held. Look hardest for nouns tied to obligations.

Who should be in the room.

MCM 080 Who to Involve

The method maps things with frontline users, supervisors, compliance staff, technology staff and analysts together. Systems engineering guidance agrees that leaving out a stakeholder often means missing needs. Never map with one side alone.

The first task.

MCM 081 Listening for Nouns

The first task is to listen for nouns and write down anything that sounds like a thing the system might hold, such as Incident Report, Location or Action Plan. Do not judge duplicates yet; sorting comes after collecting.

Seven entries on the Object Inventory, core things, zones, depth, and the pitfalls to avoid.

The inventory.

MCM 082 The Object Inventory

The Object Inventory lists each candidate thing with its name, type, short description, primary zone, importance and notes. A rough inventory everyone can see beats a precise one in one person's head.

Core and supporting.

MCM 083 Core and Supporting Things

Each thing is marked core, supporting, external or legacy only. The Eligibility Determination is core; the Notice explaining it is supporting. The method asks for three to seven core things; more usually means the list is unsorted.

Zones.

MCM 084 Zones

Zones are areas of work where things mostly live, such as intake, processing and audit. A Case appears in several, but its primary zone is where it is created and changed. Aim for three to six, in your organization's own words.

How deep to go.

MCM 085 How Deep Is Deep Enough

Field-level detail is too deep for phase one; being unable to explain a thing briefly is too shallow. Aim for one to three sentences, a zone, neighbors, and core or supporting. Field lists belong to cards and Packs.

Five pitfalls.

MCM 086 Five Mapping Pitfalls

The book names five pitfalls: a process workshop, an over-engineered map, a map shaped by old tools, forgotten compliance records, and AI defining the world. Each pulls the session away from things.

The AI pitfall.

MCM 087 Letting AI Define Your World

The fifth pitfall is accepting a tool's list of things wholesale. The book's example of an invented thing is a Customer Happiness Aura Score. Use AI as a stimulus, not an authority: compare, reject or rename, and add what it missed.

A decision is a thing.

MCM 088 A Determination Is a Thing

The book's benefits example ends with a realization: an Eligibility Determination is not a flag on a case but a thing in its own right, with its own fields, lifecycle and links. A decision stored as a flag cannot be explained or audited later.

Eight entries on doing phase one: interviews, AI, rows, the map and the workshop. The opening question is answered here.

Four interview questions.

MCM 089 Four Interview Questions

The book's four interview questions move from daily work to accountability. What do you actually touch? What are you responsible for keeping current? What would you write on paper if the system went down? And which records would an auditor ask for tomorrow? That answers this unit's question: the last question finds the things nobody touches daily, like the Policy Version and Case Snapshot, because they matter only when a decision must be proved.

Mining documents with AI.

MCM 090 Mining Documents With AI

The method uses AI to find nouns in long documents, with Course One's grounding rules. Ask for the things the text describes, using only what it supports, with uncertain items as questions. Then merge duplicates and flag ambiguous words for the workshop. Only people can say which nouns are real things.

Writing a row.

MCM 091 Writing an Inventory Row

An inventory row has six fields: the name people actually say, the type, a business-language description, the primary zone, the importance, and notes on synonyms and ambiguity. Rough is fine; a row concrete enough to argue about has done its job.

Drawing the map.

MCM 092 Drawing the Map

The map takes four steps: draw the zones, place each thing, draw arrows labelled with verbs such as becomes, produces and logs into, and highlight the core things. Focus on creation, containment and audit links; a map with too many arrows hides the ones that matter.

The book's hospital map.

MCM 093 The Hospital Map

The book's hospital map has four zones: intake, quality and risk, reporting, and compliance. An Incident Report may create a Case, which produces an Action Plan with many Tasks; the report also feeds the dashboard and logs into the audit log. The book counts a messy but recognizable picture as success.

The workshop.

MCM 094 Running the Mapping Workshop

The book's workshop runs ninety minutes or more with five to ten people: explain the purpose, brainstorm nouns, cluster into zones, place things, draw relationships, pick core things, and close. Steer away from screens and steps, park disagreements, use people's own words, and photograph the board. A disagreement about a Case is a finding, not a failure.

AI in phase one.

MCM 095 Phase One With AI, Five Steps

The book's five-step routine puts AI at both ends. Seed it with context and ask for things. Review: merge, remove inventions, add compliance records. Ask for zones. Critique and rename them. Then build the inventory and map yourselves, and only afterwards ask the tool to describe the map for stakeholders.

The book's benefits agency.

MCM 096 The Benefits Agency, Mapped

The book follows a benefits agency through five steps: collect nouns, build the inventory, name five zones, place things, and draw relationships, such as an Application becoming a Case that produces a Determination linked to a Policy Version and Snapshot. The most valuable output was the realization that the Determination is a thing.

Three constructed examples that build the library's inventory and map, and find a thing nobody had named.

The library's inventory.

MCM 097 The Library Inventory

A worked example, constructed for this course, builds the library's inventory. A noun storm with desk staff grows six things to thirteen, including the Loan Policy, which had changed without anyone updating the overdue notices. Four are core: Member, Item, Loan and Loan Policy. The inventory made visible what the work already depended on.

The library's map.

MCM 098 The Library Map

The library's constructed map has four zones in its own words: membership, circulation, collection, and accounts. Among the labelled arrows, one stands out: a Loan follows a Loan Policy. That is the link that broke last year, because nobody had drawn it. Each arrow is a dependency someone will need later.

The thing nobody named.

MCM 099 The Audit Thing We Missed

The auditor question finds a gap in the library's constructed map: fines are audited and disputed, but the Loan Policy has no versions. The team adds a Loan Policy Version and a Fine Calculation Record, both audit-critical. It took two minutes in the workshop, and would have taken a rebuild after an audit failure.

Look at what you can now do. You can say what phase one is for, where things come from, and who must be in the room. You can build an Object Inventory with zones and three to seven core things. You can avoid the mapping pitfalls, and keep AI a stimulus rather than an authority. And you can run the interviews and the workshop, ask the auditor question, and draw a map with a verb on every arrow. Unit Five writes the cards.