MWMS Meeting Intelligence And Action Extraction Framework

Parent Page: Operations Brain Canon

Brain: Operations Brain

Page Type: Framework

Status: Active

Version: v1.1

Date: 2026-06-21

Author: HeadOffice

Strategic Purpose

The MWMS Meeting Intelligence And Action Extraction Framework defines how MWMS converts meetings, calls, recorded discussions, and collaborative conversations into structured operational intelligence, assigned actions, accountable ownership, deadlines, follow-up requirements, proposal inputs where applicable, and reusable organisational evidence.

The framework prevents meetings from becoming passive conversations that depend on memory, informal notes, or individual interpretation.

Its purpose is to ensure that every meaningful meeting can produce a controlled operational result.

The framework establishes the operating sequence:

Conversation

→ Capture

→ Interpretation

→ Decision Extraction

→ Action Extraction

→ Ownership

→ Deadline

→ Validation

→ Routing

→ Follow-Up

→ Completion Evidence

→ Organisational Learning

Where a sales or client meeting may support a proposal, the framework may also produce:

→ Proposal Input Pack

→ Commercial Review Handoff

The meeting-intelligence system must stop before commercial approval.

Core Principle

A meeting is not complete when the conversation ends.

A meeting is complete only when:

  • material decisions have been captured
  • required actions have been identified
  • every action has an owner
  • every time-sensitive action has a deadline
  • unclear commitments have been resolved or escalated
  • outputs have been routed to the correct operating destination
  • completion can later be verified
  • proposal-relevant inputs have been separated from approved commercial commitments where applicable

Meeting intelligence is therefore not merely summarisation.

It is the controlled conversion of human conversation into operational execution and governed downstream inputs.

Framework Scope

This framework applies to:

  • internal MWMS meetings
  • HeadOffice planning meetings
  • Brain-to-Brain coordination meetings
  • employee meetings
  • AI employee supervision sessions
  • project reviews
  • client onboarding calls
  • client service meetings
  • sales calls
  • partnership discussions
  • supplier discussions
  • strategy sessions
  • performance reviews
  • training sessions
  • operational handoffs
  • problem-solving calls
  • recorded voice instructions
  • discovery meetings
  • proposal-scoping meetings
  • commercial review meetings

This framework governs:

  • meeting capture
  • transcript preparation
  • discussion extraction
  • decision extraction
  • action extraction
  • ownership
  • deadlines
  • risk and exception extraction
  • follow-up
  • routing
  • completion evidence
  • proposal-input extraction

This framework does not independently govern:

  • proposal pricing
  • commercial approval
  • legal contract approval
  • scope authorisation
  • discount approval
  • guarantee approval
  • client delivery commitments

Those matters are governed by the MWMS Client Proposal Generation And Commercial Approval Framework and the relevant Sales, Finance, Legal, Compliance, Risk, and HeadOffice authorities.

Meeting Intelligence Architecture

The framework operates across the following layers:

Layer One: Raw Meeting Evidence

Layer Two: Normalised Meeting Record

Layer Three: Discussion Intelligence

Layer Four: Decision Intelligence

Layer Five: Action Intelligence

Layer Six: Risk And Exception Intelligence

Layer Seven: Follow-Up Intelligence

Layer Eight: Organisational Learning Intelligence

Layer Nine: Proposal Input Intelligence Where Applicable

Layer One: Raw Meeting Evidence

This layer contains the original meeting evidence.

It may include:

  • audio recording
  • video recording
  • transcript
  • speaker-attributed transcript
  • written meeting notes
  • chat messages
  • attached documents
  • screen-shared information
  • participant submissions

Raw evidence must be preserved where retention is permitted and operationally justified.

The system must not silently replace raw evidence with an AI-generated summary.

Layer Two: Normalised Meeting Record

The raw conversation is converted into a consistent meeting record.

The normalised record should identify:

  • meeting title
  • meeting type
  • date
  • start and finish time where available
  • participants
  • participant roles
  • source location
  • recording or transcript reference
  • meeting purpose
  • agenda where available
  • processing date
  • responsible Brain or business area

Speaker attribution should be preserved where technically reliable.

Uncertain speaker attribution must be labelled as uncertain.

Layer Three: Discussion Intelligence

The system identifies the substantive subjects discussed.

Discussion intelligence may include:

  • topic
  • context
  • facts presented
  • positions expressed
  • evidence referenced
  • options considered
  • objections
  • constraints
  • assumptions
  • unresolved questions

This layer describes what was discussed without yet converting everything into an action or approved commitment.

Layer Four: Decision Intelligence

The system identifies decisions that were actually made.

Each decision record should contain:

  • decision statement
  • decision owner or authority
  • date of decision
  • affected Brain, project, client, or system
  • supporting rationale where available
  • rejected alternatives where relevant
  • dependencies
  • conditions
  • review date where required
  • source evidence reference

A suggestion must not be recorded as a decision.

A preference must not be recorded as an approval.

An unresolved discussion must not be recorded as an agreed direction.

A proposal idea must not be recorded as approved scope.

A requested price must not be recorded as approved pricing.

A desired completion date must not be recorded as a guaranteed deadline.

Layer Five: Action Intelligence

The system identifies work that must occur because of the meeting.

Each action should contain:

  • action statement
  • responsible owner
  • supporting participants where relevant
  • due date
  • priority
  • destination
  • dependency
  • completion criteria
  • required approval
  • escalation condition
  • source decision or discussion
  • current status

Layer Six: Risk And Exception Intelligence

The system identifies matters requiring caution, clarification, or escalation.

This may include:

  • unclear authority
  • disputed decision
  • missing owner
  • missing deadline
  • confidentiality
  • privacy
  • regulatory risk
  • client dissatisfaction
  • financial commitment
  • scope uncertainty
  • pricing uncertainty
  • legal terms
  • unsupported claim
  • sensitive employee information
  • dependency risk
  • delivery risk
  • transcript uncertainty

Layer Seven: Follow-Up Intelligence

The system identifies what follow-up communication or confirmation is required.

Follow-up may include:

  • participant summary
  • action confirmation
  • client email
  • owner reminder
  • approval request
  • deadline reminder
  • dependency request
  • escalation notice
  • next meeting booking
  • missing-information request
  • proposal clarification request
  • commercial review request

Follow-up messages should be generated from validated meeting intelligence, not from unreviewed assumptions.

Layer Eight: Organisational Learning Intelligence

Meeting outputs may identify reusable learning such as:

  • repeated bottlenecks
  • recurring objections
  • process failures
  • training needs
  • unclear responsibilities
  • repeated client questions
  • common delivery risks
  • recurring proposal ambiguities
  • repeated scope-change causes

Learning intelligence must be routed through the appropriate MWMS evidence and knowledge-governance process before becoming Canon.

Layer Nine: Proposal Input Intelligence Where Applicable

Sales, discovery, client, partnership, and solution-design meetings may contain information that should support a commercial proposal.

The meeting-intelligence system may extract:

  • current situation
  • client problem
  • desired outcome
  • confirmed requirements
  • probable requirements
  • unconfirmed requirements
  • stated constraints
  • technical constraints
  • business constraints
  • budget indicators
  • requested timing
  • stakeholders
  • decision process
  • approval process
  • must-have requirements
  • nice-to-have requirements
  • risks
  • objections
  • open questions
  • proposed next step

This layer produces a Proposal Input Pack.

It does not produce an approved proposal.

Proposal Input Pack

Client:

Company:

Opportunity:

Meeting ID:

Meeting Date:

Participants:

Source Records:

Current Situation:

Problem:

Desired Outcome:

Confirmed Requirements:

Probable Requirements:

Unconfirmed Requirements:

Constraints:

Stakeholders:

Budget Indicators:

Requested Timing:

Decision Process:

Approval Process:

Risks:

Objections:

Open Questions:

Recommended Next Step:

Transcript Confidence:

Human Review Required:

Proposal Input Status:

Proposal Input Status Values

Ready For Commercial Review

Incomplete

Clarification Required

Conflicting Evidence

Low Confidence

Not Suitable For Proposal

Rule

A Proposal Input Pack is a governed handoff artifact.

It is not authority to:

  • approve scope
  • approve deliverables
  • approve price
  • apply a discount
  • approve payment terms
  • promise a deadline
  • promise a result
  • create a guarantee
  • send a proposal
  • begin delivery

The dedicated MWMS Client Proposal Generation And Commercial Approval Framework governs all later commercial steps.

Meeting Processing Lifecycle

Stage One: Capture Authority

Before a meeting is recorded or processed, the system must establish whether capture is authorised.

Capture authority may arise from:

  • participant consent
  • contractual permission
  • internal policy
  • approved meeting platform settings
  • client agreement
  • employment policy
  • legal authority

Unauthorised recording or processing must not occur.

Stage Two: Source Capture

The authorised meeting source is captured.

The system should preserve:

  • original source
  • source timestamp
  • participant details
  • recording reference
  • transcript reference
  • source integrity
  • processing status

Stage Three: Transcript Preparation

Where a transcript is used, it should be prepared for analysis.

Preparation may include:

  • removing transcription artefacts
  • preserving speaker labels
  • correcting obvious formatting errors
  • removing irrelevant system messages
  • separating conversational turns
  • retaining meaningful timestamps
  • marking uncertain wording

Preparation must not rewrite the substance of the conversation.

Stage Four: Meeting Classification

The meeting is classified before specialised extraction occurs.

Possible meeting classes include:

  • internal operations meeting
  • client onboarding meeting
  • client service meeting
  • sales call
  • discovery call
  • proposal-scoping meeting
  • commercial review meeting
  • project review
  • performance meeting
  • strategic planning meeting
  • partnership meeting
  • training session
  • incident review
  • handoff meeting
  • approval meeting

Meeting classification determines which extraction rules and checklists apply.

Stage Five: Intelligence Extraction

The system extracts:

  • summary
  • topics
  • facts
  • decisions
  • actions
  • owners
  • deadlines
  • risks
  • questions
  • dependencies
  • follow-up requirements
  • relevant performance evidence
  • proposal inputs where applicable

Extraction should use a structured output format rather than unrestricted prose where operational routing is required.

Stage Six: Ambiguity Detection

The system checks for incomplete operational or commercial information.

Examples include:

  • “Someone should handle this.”
  • “We will do that soon.”
  • “Let us revisit it later.”
  • “M can probably take care of it.”
  • “We agreed to change it.”
  • “We can probably include that.”
  • “The price should be around this.”
  • “They want it next month.”

These statements may imply work or commercial direction, but they lack sufficient execution or approval detail.

The system should identify:

  • unknown owner
  • unknown deadline
  • unclear action
  • unclear authority
  • unclear destination
  • unclear completion condition
  • unclear scope
  • unclear price
  • unclear commercial term
  • unclear proposal requirement

Ambiguous items must be resolved, held for review, or escalated.

They must not be silently converted into confident tasks or commercial commitments.

Stage Seven: Human Validation

Material outputs should be reviewed before they create irreversible operational consequences.

Validation may confirm:

  • summary accuracy
  • decision accuracy
  • action accuracy
  • ownership
  • deadline
  • priority
  • routing destination
  • confidentiality
  • required approval
  • proposal-input accuracy
  • requirement classification
  • transcript confidence

Low-risk routine actions may use approved automated routing where governance permits.

High-impact decisions, client commitments, financial commitments, employee matters, policy changes, scope changes, pricing, discounts, guarantees, and delivery commitments require human validation.

Stage Eight: Operational Routing

Validated outputs are routed to the correct destination.

Possible destinations include:

  • task system
  • project record
  • client record
  • CRM
  • Brain queue
  • HeadOffice review
  • employee task list
  • approval queue
  • risk register
  • decision log
  • follow-up communication draft
  • meeting archive
  • lessons learned system
  • Canon review queue
  • Proposal Input Pack
  • commercial review queue

Routing must follow MWMS Brain authority and destination rules.

A Proposal Input Pack must route to the MWMS Client Proposal Generation And Commercial Approval Framework or its approved operational destination.

Stage Nine: Follow-Up

The system determines whether follow-up is required.

Follow-up may include:

  • participant summary
  • action confirmation
  • client email
  • owner reminder
  • approval request
  • deadline reminder
  • dependency request
  • escalation notice
  • next meeting booking
  • missing-information request
  • proposal clarification
  • commercial approval request

Follow-up messages should be generated from validated meeting intelligence, not from unreviewed assumptions.

Stage Ten: Completion Verification

Actions remain open until completion evidence is available.

Completion evidence may include:

  • task marked complete
  • document produced
  • approval recorded
  • client response received
  • system change verified
  • payment confirmed
  • handoff accepted
  • deliverable reviewed
  • issue resolved
  • Proposal Input Pack accepted for commercial review

A completed conversation does not constitute a completed action.

A completed Proposal Input Pack does not constitute an approved proposal.

Decision Extraction Standard

A decision must satisfy the following conditions before it is recorded as confirmed:

  • a choice or direction was clearly established
  • the relevant authority made or accepted the decision
  • the decision applies to an identifiable matter
  • the wording is sufficiently clear to act upon
  • material conditions are captured
  • the source can be traced

Decision classifications include:

Confirmed Decision

A clear decision was made and can be acted upon.

Conditional Decision

A decision was made subject to a condition.

Provisional Direction

A temporary direction was given but requires later confirmation.

Recommendation

An option was proposed but not approved.

Deferred Decision

The matter was intentionally postponed.

Disputed Decision

Participants do not agree on what was decided.

No Decision

The matter was discussed without reaching a decision.

Commercial Discussion

A possible commercial matter was discussed but has not passed commercial approval.

Rule

Commercial Discussion must not be represented as Confirmed Decision unless the correct commercial authority explicitly approved it.

Action Extraction Standard

A valid operational action should answer:

  • What must be done?
  • Who owns it?
  • When is it due?
  • Why is it required?
  • What decision or request created it?
  • What does completion look like?
  • What dependencies exist?
  • Where must the result be delivered?

Where one of these elements is absent, the action should be marked incomplete rather than fabricated.

Action Ownership Rule

Every action must have one accountable owner.

Multiple contributors may support an action, but shared ownership must not remove accountability.

The accountable owner is responsible for:

  • accepting the action
  • clarifying uncertainty
  • coordinating contributors
  • managing dependencies
  • meeting the deadline
  • providing completion evidence
  • escalating blockers

Actions without an owner must enter an unresolved-action queue.

Deadline Rule

The system must distinguish between:

  • explicit deadline
  • inferred deadline
  • requested timeframe
  • target date
  • review date
  • no deadline provided

Inferred deadlines must be clearly labelled and require confirmation.

The system must not convert words such as “soon,” “later,” or “when possible” into arbitrary dates without approved scheduling logic.

Commercial Timing Rule

The system must also distinguish between:

  • client-requested date
  • preferred date
  • target date
  • estimated date
  • dependent date
  • fixed approved commitment

A client-requested or preferred date must not be represented as a fixed delivery commitment.

Completion Criteria Rule

An action should not be considered execution-ready unless its completion condition is understandable.

Weak completion condition:

“Review the client system.”

Strong completion condition:

“Review the client system against the approved checklist, document all failures, assign severity, and submit the review to HeadOffice.”

Where completion criteria cannot be determined, the action must be clarified before execution.

Meeting Checklist Evaluation

Different meeting classes may use approved checklists.

A checklist may evaluate whether required subjects or actions occurred.

Examples include:

  • Was the meeting purpose stated?
  • Were required participants present?
  • Was the client’s objective confirmed?
  • Were unresolved issues identified?
  • Were decisions explicitly confirmed?
  • Were owners assigned?
  • Were deadlines agreed?
  • Were risks discussed?
  • Was the next step explained?
  • Was approval requested where required?
  • Were proposal-relevant requirements separated into confirmed and unconfirmed?
  • Were requested dates distinguished from approved commitments?
  • Was commercial review identified where required?

Checklist evaluation may produce:

  • completed
  • partially completed
  • not completed
  • not applicable
  • unclear from evidence

Checklist results are evidence for review and improvement.

They are not automatically employee performance judgments.

Performance Evidence Rule

Meeting data may support performance analysis when appropriately governed.

Possible evidence includes:

  • required questions asked
  • commitments completed
  • deadlines established
  • customer concerns addressed
  • agreed process followed
  • next steps confirmed
  • communication clarity
  • meeting participation
  • repeated omissions

Performance evidence must:

  • use an approved role-specific standard
  • preserve context
  • distinguish fact from interpretation
  • permit human review
  • avoid automatic disciplinary conclusions
  • avoid scoring people against irrelevant meeting types
  • protect sensitive employee information

Sales Call Intelligence

For sales calls, the framework may extract:

  • prospect goals
  • pain points
  • current state
  • desired outcome
  • urgency
  • budget indicators
  • authority indicators
  • objections
  • decision process
  • promised follow-up
  • offer discussed
  • proof presented
  • agreed next step
  • deal risk
  • proposal-input readiness

Sales-call intelligence must route into the appropriate Sales Brain or CRM workflow.

It must not remain isolated inside a transcript archive.

Where a proposal may be appropriate, the system should generate a Proposal Input Pack.

It must not independently create or approve:

  • scope
  • deliverables
  • pricing
  • discounts
  • payment terms
  • guarantees
  • deadlines
  • contract wording
  • send authority

Client Meeting Intelligence

For client meetings, the framework may extract:

  • client requests
  • scope changes
  • dissatisfaction
  • desired outcomes
  • agreed deliverables
  • deadlines
  • approvals
  • rejected work
  • new opportunities
  • risk signals
  • relationship context
  • client-specific facts

Client requests and possible scope changes must be classified.

Possible classifications:

Confirmed In-Scope Action

Possible Scope Change

Commercial Review Required

Out Of Scope

Unclear

Rule

A client request raised in conversation must not silently become included scope.

Proposal Handoff Boundary

Meeting Intelligence may produce:

  • source-linked requirements
  • confirmed statements
  • probable requirements
  • unconfirmed requirements
  • constraints
  • budget indicators
  • requested timing
  • stakeholders
  • objections
  • open questions
  • recommended next step

Meeting Intelligence may not approve:

  • final scope
  • final deliverables
  • final timeline
  • final price
  • discounts
  • payment schedule
  • commercial claims
  • guarantees
  • legal terms
  • proposal validity
  • external proposal delivery

The handoff destination is:

MWMS Client Proposal Generation And Commercial Approval Framework

Proposal Handoff Record

Meeting ID:

Opportunity ID:

Client:

Proposal Input Pack Version:

Source Confidence:

Commercial Review Required:

Open Questions:

Required Approvers:

Destination:

Handoff Status:

Handoff Status Values

Not Required

Preparing

Ready For Review

Clarification Required

Accepted By Commercial Review

Rejected By Commercial Review

Superseded

AI Employee Meeting Use

AI employees may assist with:

  • transcription
  • summarisation
  • decision extraction
  • action extraction
  • checklist evaluation
  • routing recommendations
  • reminder generation
  • follow-up drafting
  • completion checking
  • proposal-input extraction
  • requirement classification
  • ambiguity detection

AI employees must not independently:

  • create policy from conversation
  • assign disciplinary consequences
  • approve financial commitments
  • change Canon
  • create unauthorised client commitments
  • invent missing owners or deadlines
  • expose confidential meeting information
  • route sensitive employee information to general systems
  • approve proposal scope
  • approve pricing
  • approve discounts
  • approve payment terms
  • promise delivery dates
  • create guarantees
  • send proposals

Source Traceability

Every material meeting output should retain a reference to its source.

Traceability may include:

  • meeting identifier
  • transcript identifier
  • recording identifier
  • timestamp
  • speaker reference
  • source excerpt
  • processing version
  • validation status

A material decision, commitment, or proposal input should be traceable to the part of the conversation that supports it.

Confidence And Uncertainty

The system must distinguish:

  • explicit statement
  • strong inference
  • weak inference
  • unresolved ambiguity
  • conflicting evidence

Confidence must not be used to disguise missing information.

When confidence is low, the system should request clarification or route the item for review.

Proposal Input Confidence

High

Material proposal inputs are explicit, attributable, and internally consistent.

Medium

Most inputs are usable but some require confirmation.

Low

Material ambiguity, missing context, or transcript quality issues exist.

Unusable

The meeting evidence cannot safely support proposal drafting.

Rule

Low or unusable proposal-input confidence must block automatic proposal drafting.

Privacy And Confidentiality

Meeting intelligence may contain highly sensitive information.

The system must apply:

  • least-privilege access
  • participant consent requirements
  • client isolation
  • employee information controls
  • retention limits
  • access logging
  • secure storage
  • restricted routing
  • deletion requirements where applicable
  • redaction where full transcript access is unnecessary

Sensitive meeting content must not be placed into general-purpose knowledge systems without explicit authority.

Retention Standard

Not every meeting requires permanent storage.

Retention should consider:

  • legal requirement
  • contractual requirement
  • operational value
  • sensitivity
  • dispute risk
  • client relationship value
  • training value
  • decision significance
  • Canon relevance
  • proposal evidence value

Possible retention outcomes include:

  • temporary processing only
  • short-term operational record
  • client record
  • project record
  • decision archive
  • restricted employee record
  • commercial evidence record

Failure Handling

Possible failures include:

  • recording unavailable
  • transcript incomplete
  • wrong meeting matched
  • speaker attribution failure
  • confidential information exposed
  • decision misclassified
  • action fabricated
  • owner missing
  • deadline fabricated
  • routing failure
  • failed task creation
  • failed follow-up
  • action completed without evidence
  • proposal input misclassified
  • commercial commitment inferred
  • requested date represented as fixed
  • possible scope treated as approved scope

Failure must not be hidden behind a polished summary.

Minimum Meeting Output

A standard meeting output should contain:

Meeting Identity

  • meeting title
  • meeting type
  • date
  • participants
  • purpose

Executive Summary

  • concise explanation of what occurred

Confirmed Decisions

  • decision
  • authority
  • affected area
  • conditions

Action Register

  • action
  • owner
  • due date
  • priority
  • completion criteria

Open Questions

  • unresolved matter
  • required owner
  • required resolution date

Risks And Dependencies

  • risk or dependency
  • affected action
  • required response

Follow-Up Requirements

  • communication
  • recipient
  • timing
  • approval requirement

Proposal Input Pack Where Applicable

  • current situation
  • problem
  • desired outcome
  • confirmed requirements
  • unconfirmed requirements
  • requested timing
  • open questions
  • transcript confidence
  • commercial review status

Evidence

  • transcript or recording reference
  • validation status

Operational Status

  • routed
  • awaiting review
  • awaiting clarification
  • active
  • completed
  • escalated

Quality Standard

A meeting intelligence output is acceptable only when it is:

  • faithful to the source
  • operationally useful
  • clear about uncertainty
  • structured for routing
  • explicit about ownership
  • explicit about deadlines
  • traceable
  • appropriately confidential
  • governed by human validation where required
  • free from invented commitments
  • commercially bounded where proposal inputs are involved

Success Measures

The framework should improve:

  • percentage of meetings with captured decisions
  • percentage of actions with owners
  • percentage of actions with deadlines
  • percentage of actions with completion criteria
  • follow-up speed
  • action completion rate
  • reduction in forgotten commitments
  • reduction in ownership ambiguity
  • reduction in repeated discussion
  • meeting-to-execution time
  • cross-Brain handoff reliability
  • client follow-up reliability
  • visibility of recurring blockers
  • percentage of proposal-input packs with source traceability
  • percentage of proposal-input packs with clear confidence status
  • reduction in unauthorised commercial commitments

Success must not be measured by the number of transcripts created.

Governance Rules

  1. A transcript is evidence, not authority.
  2. A summary must not replace the source where material decisions are involved.
  3. A suggestion must not be represented as a decision.
  4. An inferred action must not be presented as an explicit commitment.
  5. Every operational action requires an accountable owner.
  6. Every time-sensitive action requires a deadline or clarification status.
  7. Sensitive meeting intelligence must be access controlled.
  8. A Proposal Input Pack is not an approved proposal.
  9. Requested scope must not be represented as approved scope.
  10. Requested timing must not be represented as a guaranteed delivery date.
  11. Budget discussion must not be represented as approved pricing.
  12. AI may extract proposal inputs but may not approve commercial commitments.
  13. Commercial outputs must route through the MWMS Client Proposal Generation And Commercial Approval Framework.

Framework Outcome

When this framework is operating correctly, MWMS meetings become controlled sources of execution intelligence and governed commercial inputs.

Conversations no longer disappear into recordings, informal notes, or participant memory.

Decisions become visible.

Actions become owned.

Deadlines become explicit.

Risks become traceable.

Follow-up becomes systematic.

Proposal inputs become structured.

Commercial ambiguity becomes visible.

Completion becomes verifiable.

Learning becomes reusable.

The framework therefore converts meetings from temporary communication events into governed operational inputs for MWMS execution, accountability, coordination, client delivery, organisational learning, and commercial review.

MWMS System Change Log

Version: v1.1

Date: 2026-06-21

Author: HeadOffice

Change

Updated the MWMS Meeting Intelligence And Action Extraction Framework from v1.0 to v1.1 following creation of the MWMS Client Proposal Generation And Commercial Approval Framework.

Added a dedicated Proposal Input Intelligence layer.

Added the Proposal Input Pack containing:

  • client and opportunity identity
  • meeting and source records
  • current situation
  • problem
  • desired outcome
  • confirmed requirements
  • probable requirements
  • unconfirmed requirements
  • constraints
  • stakeholders
  • budget indicators
  • requested timing
  • decision process
  • approval process
  • risks
  • objections
  • open questions
  • recommended next step
  • transcript confidence
  • human review requirement

Added proposal-input statuses:

  • Ready For Commercial Review
  • Incomplete
  • Clarification Required
  • Conflicting Evidence
  • Low Confidence
  • Not Suitable For Proposal

Added a Proposal Handoff Boundary defining that Meeting Intelligence may extract and route source-linked commercial inputs but may not independently approve:

  • scope
  • deliverables
  • pricing
  • discounts
  • payment terms
  • guarantees
  • deadlines
  • contract wording
  • proposal validity
  • external delivery

Added Proposal Handoff Record and handoff statuses.

Expanded:

  • meeting scope
  • meeting classification
  • ambiguity detection
  • human validation
  • operational routing
  • follow-up
  • completion verification
  • sales call intelligence
  • client meeting intelligence
  • AI employee permissions
  • source traceability
  • confidence handling
  • failure handling
  • minimum meeting output
  • success measures
  • governance rules

Added explicit doctrine that:

  • a Proposal Input Pack is not an approved proposal
  • a client request does not automatically become included scope
  • a requested date does not automatically become a fixed delivery commitment
  • budget discussion does not create approved pricing
  • AI may extract proposal inputs but may not approve commercial commitments

Change Impact Declaration

This update creates a formal controlled boundary between Operations Brain meeting intelligence and Sales Brain commercial proposal governance.

Operations Brain remains responsible for:

  • meeting capture
  • transcript preparation
  • discussion extraction
  • decision extraction
  • action extraction
  • proposal-input extraction
  • ambiguity detection
  • routing

Sales Brain, Finance Brain, Risk Brain, Compliance Brain, and authorised commercial reviewers remain responsible for:

  • proposal scope
  • deliverables
  • pricing
  • payment terms
  • claims
  • timeline commitments
  • commercial approval
  • external delivery

The update does not authorise automatic proposal generation or sending.

The dedicated downstream authority is:

  • MWMS Client Proposal Generation And Commercial Approval Framework

Pages Created

  • None

Pages Updated

  • MWMS Meeting Intelligence And Action Extraction Framework updated from v1.0 to v1.1

Pages Deprecated

  • None

Standalone Pages Not Created

The following standalone pages were not created because their durable intelligence is governed within this updated framework or the dedicated proposal framework:

  • MWMS Meeting To Proposal Handoff Framework
  • MWMS Proposal Input Pack Standard
  • MWMS Transcript To Proposal Input Framework
  • MWMS Sales Call Proposal Extraction Framework
  • MWMS Commercial Meeting Handoff Protocol

Registries Requiring Update

  • MCR Page Registry
  • Operations Brain Page Registry
  • Sales Brain Page Registry where the proposal handoff is referenced
  • MCR Copy Map where the Operations Brain framework copy is recorded
  • MWMS Course Absorption Decision Registry

Canon Version Update Required

No immediate Operations Brain Canon version change is required unless the current Canon records framework versions or meeting-to-proposal handoff rules.

The next scheduled Operations Brain Canon alignment should recognise:

  • Proposal Input Intelligence
  • Proposal Input Pack
  • Proposal Handoff Boundary
  • commercial-review routing

Change Log Entry Required

Yes.

The v1.1 update must be recorded in:

  • MWMS System Change Log
  • MCR Page Registry change history where applicable
  • Operations Brain Page Registry change history where applicable
  • Sales Brain Page Registry change history where applicable
  • MWMS Course Absorption Decision Registry

Strategic Absorption Result

The AI Automations by Jack proposal-generation material has been connected to the existing MWMS meeting-intelligence framework without allowing meeting analysis to become uncontrolled commercial authority.

The absorption preserves:

  • transcript-based requirement extraction
  • problem and desired-outcome extraction
  • requested timing
  • budget indicators
  • stakeholder information
  • objections
  • follow-up
  • source traceability

The absorption rejects:

  • turning informal discussion into approved scope
  • turning requested timing into guaranteed deadlines
  • turning budget discussion into approved pricing
  • allowing AI to approve discounts or payment terms
  • automatically generating and sending proposals from a meeting transcript

The resulting v1.1 framework establishes that meeting intelligence may create a source-linked Proposal Input Pack, but all commercial commitments must pass through the MWMS Client Proposal Generation And Commercial Approval Framework.

END OF FULL FILE OUTPUT