Services
Fragment Practice helps organizations clarify ambiguous issues around AI adoption, internal rules, security expectations, external services, and technology risk, then turn them into material that can be used for decisions, stakeholder explanation, review, operation, and handoff.
Fragment Practice works close to important decisions, with a defined scope, agreed cadence, and usable output. Engagements can start small — from an initial structuring session — and ongoing support runs as scoped advisory with an agreed theme and role.
Service model
AI, governance, and security work often mixes use cases, requirements, roles, and responsibility boundaries. Fragment Practice separates those issues into material that can be used for explanation, review, operation, and handoff.
AI and security initiatives, ready to decide, explain, and hand off
Where initiatives get stuck
What Fragment Practice does
A state ready for decision and execution
Examples: decision memo, use-case and control structure, responsibility map, review and approval flow, requirements, roadmap, and handoff material for implementation and operations.
Areas of support
Depending on the current situation, these areas are combined as needed — useful when the issue is real, but the decision criteria, review points, internal roles, stakeholder explanation, or next-step material are still unclear.
Area of support
Turn AI governance direction, use rules, and decision criteria into material teams can actually use — what can be entered, how output is used, who reviews, what to record, and where responsibility sits.
Area of support
Separate which judgments people keep from what is delegated to AI, automation, or an external party — then set who acts, who reviews, on what condition work stops and returns for consultation, and how it is recorded and handed off.
Area of support
When a customer sends a security questionnaire, or a new external service is up for adoption, Fragment Practice translates security requirements, technology risks, evidence needs, stakeholder concerns, and practical constraints into reviewable next-step material.
Area of support
Settle who judges an incident, who is told, at what point it escalates, and what gets recorded — so response does not depend on who happens to be on duty.
Area of support
Read an external requirement against what the organisation already does, and turn the gap into priorities, owners, and review points it can act on.
Ways to engage
The engagement shape does not need to be decided before the first conversation; it can be defined together after reviewing the situation. Start with an initial structuring session, use a decision-material sprint when material is needed soon, or continue with scoped advisory support when review and interpretation needs keep returning.
See what to decide first
Clarifies the issue, decision order, review points, and next actions when the situation is still too mixed to scope cleanly.
Price guide
From ¥300,000
Excl. tax. Typically 1–2 sessions, including advance review of existing material.
Best for
When the issue exists, but the decision point, review need, role boundary, or next move is still unclear.
Includes
Leaves behind
Ready before the decision
Assembles the issues, responsibility boundaries, review points, and outputs that a dated decision, rule update, third-party review, or handoff needs.
Price guide
From ¥1,200,000
Excl. tax. Usually 2–6 weeks. Scoped by output, cadence, and involvement.
Best for
When a meeting, report, decision, review, rule update, or handoff needs usable material soon.
Includes
Leaves behind
Keep one theme under review
Provides continuing expert judgment and review on a defined theme, keeping decision material current at the points that matter.
Price guide
From ¥600,000/month
Excl. tax. Designed by theme, cadence, review scope and volume, response expectations, outputs, and responsibility boundaries.
Best for
When recurring external advice is needed without a resident or all-in role, and scope and response expectations should stay bounded.
Includes
Leaves behind
These are starting points (excl. tax) for new engagements, not fixed packages. Final scope and fees are designed around the theme, number of stakeholders, outputs, meeting cadence, and asynchronous response. Existing agreements are unaffected.
Representative outputs
Outputs are designed to be usable across all of these areas: for management and stakeholder explanation, practical decision-making and operation, ongoing updates, and handoff to the next phase.
Material for explaining current state, decision points, risks, options, and next actions.
Contains
Helps answer
Typical questions
Often used before management discussion, stakeholder explanation, planning, rollout, or handoff.
Material for clarifying AI use cases, input/output boundaries, human review, approval logic, records, and responsibility.
Contains
Helps answer
Typical questions
Often used when AI adoption needs to become reviewable, explainable, and workable.
Material for clarifying what AI, people, internal teams, external advisors, or vendors support, review, decide, or hand off.
Contains
Helps answer
Typical questions
Often used when AI adoption, business efficiency, or external support needs to become a practical operating model.
Material for connecting internal rules, external requirements, guidelines, controls, review points, responsibility, and update cycles.
Contains
Helps answer
Typical questions
Often used when rules, requirements, guidelines, or control expectations need to become internal discussion and operating material.
An outside read of an existing policy, rule set, risk assessment, vendor proposal, or design.
Contains
Helps answer
Typical questions
A review, not assurance. It does not certify, audit, or substitute for legal judgement, and it states its own limits and what it did not assess.
A way to engage
Sometimes the policy and the documents exist, and the open question is whether they hold. An independent review says what holds, what does not, what it rests on, what risk remains, and what to decide next. It applies to any of these areas.
Typical subjects
What this is not
This is not an audit, assurance, certification, or a substitute for legal judgement. It returns findings and open questions, not a verdict, a score, or a maturity level — and it states what was not assessed.
Where this fits
The work centres on making issues easier to decide, review, explain, operate, and hand off. For ongoing support, the cadence, review scope, coverage, and response expectations are agreed before the work starts.
Good fit
The issue still needs clearer criteria, review logic, role design, responsibility boundaries, or management-facing material.
Good fit
Useful when AI adoption also involves data, routines, governance, education, internal rules, and stakeholder explanation.
Good fit
Useful when internal owners need to interpret requirements, update rules, define review points, or explain what should change.
Good fit
Useful when ongoing support is needed, but the cadence, review volume, response expectations, and responsibility boundaries can be defined.
Better handled another way
Audit assurance, certification, legal judgment, technical verification, and day-to-day progress or task management sit with other specialists or teams. This work covers decision material, review points, role / operating design, responsibility boundaries, and stakeholder explanation.
What you receive, and when
The first conversation confirms what is moving, who needs to decide or review, what feels unclear, and what material needs to remain — and from that, the support shape and the outputs. From early on you hold a written view that separates what is established, what is still a working hypothesis, what is unknown, and what has to be decided.
At the first conversation
Describe what is moving, who needs to decide or review, what feels mixed or unclear, and whether the need is one-time or recurring.
At the first conversation
Confirm whether the issue fits an initial structuring session, decision-material sprint, scoped advisory support, product kit, or another path.
At the first conversation
Clarify outputs, cadence, review volume, role, response expectations, and limits so the work stays a defined, bounded engagement — including the conditions for returning a decision, handling exceptions, and escalating.
Once the work starts
The work does not wait until everything is known. From the material received and the conversations so far, the current understanding is written down early — what is established, what is still a working hypothesis, what is unknown, and what has to be decided and by whom. It is not a finished recommendation; it is there to align how each side reads the situation.
Once the work starts
There is no need to go through it item by item. Tell us where it reads differently to you, or where you want a closer look, and the next questions are narrowed from there. The question you brought is looked at the same way rather than taken as settled. What comes out of that is added to the existing structuring rather than restarted, and what remains is material usable for explanation and for the decision.
Next step
Use an Initial structuring session when the issue needs its first clear shape. Use a Decision-material sprint when a report, meeting, rule update, or handoff needs usable material soon. Use Scoped advisory support when recurring review and interpretation are needed without replacing internal owners.
