AI

CSP on AINPX Future Vision

CSP on AINPX Future Vision

From fragmented workspaces to a personalized AI operating environment

A role-agnostic, context-aware platform that anticipates customer needs, coordinates work across roles, and assembles the right experience for every customer moment.

Future-state strategyResearch-informedDirectional visionLiving documentLast updated 2026-08-03

Today

Users navigate fixed, role-based experiences and manually assemble customer context across disconnected tools.

Future state

CSP continuously understands customers, workflows, roles, and priorities, then assembles the right context, actions, and collaboration around the moment.

Outcome

Users move from finding information to making informed decisions and advancing customer outcomes.

Horizon

Research confidence

Section 01

Strategy at a glance

The full argument in one screen: the problem, the opportunity, the anchor, and the foundation.

Strategic problem

Context

CSP has grown feature by feature and role by role.

New tabsNew dashboardsNew scorecardsNew tablesNew role workspacesNew disconnected agentsMore navigationGreater maintenance cost

The accumulated experience places the burden on users to search, interpret, validate, and decide what matters.

Validated

Strategic opportunity

Direction

AINPX creates an opportunity to redesign the underlying model rather than add AI to the current interface.

Understand customer contextWatch for meaningful changePredict emerging needsAssemble relevant informationRecommend next actionsCoordinate roles and workflowsMaintain durable account memoryPreserve human judgment and accountability
Directional evidence

Primary user anchor

Scope

Use the CSM as the initial point of view. The model extends beyond the CSM.

Customer outcomesAdoptionValue realizationSupportRiskAccount teamsExecutive relationshipsCommunicationInternal coordination
Directional evidence

Primary scenario

Scope

Use a customer risk or escalation as the primary demonstration scenario.

Fragmented informationCross-role dependenciesCustomer urgencyTrust requirementsOwnership ambiguityWorkflow coordinationHuman judgmentCustomer communicationResolution and follow-through
Directional evidence

Experience foundation

Architecture
Shared context graphAmbient watchingDynamic assemblyExplainable memoryGuided actionCross-role orchestration
Design hypothesis

Awareness model

Architecture
System AwarenessWorkflow AwarenessRole Awareness
Directional evidence

Interaction model

Experience

Hybrid is the dominant future-state pattern.

ConversationalHybrid (dominant)UI-driven
Directional evidence

Delivery horizons

Sequencing
Now — foundationNext — assisted workflowsNorth Star — assembled experiences
Design hypothesis

Section 02

Why CSP must change

Six structural problems that no additional dashboard, tab, or agent can solve.

Fixed experience growth

CSP expands through tabs, scorecards, tables, columns, and persona-specific workspaces as new needs emerge.

Impact

Navigation and cognitive load increase for users, while product and engineering inherit additional surfaces to maintain.

Validated

Fragmented context

Customer and operational context is spread across CSP, support systems, project systems, CRM, email, Teams, documents, dashboards, spreadsheets, and individual knowledge.

Impact

Users repeatedly assemble the same information before they can act confidently.

Validated

Reactive monitoring

Users must continually check dashboards, account records, support cases, project data, and health signals to determine whether intervention is needed.

Impact

Issues are often discovered after the customer raises them or after the situation has become urgent.

Validated

Incomplete handoffs

Context is lost when work moves across CSMs, Support, SAMs, PAs, TSMs, EMs, RMs, Sales, partners, and leadership.

Impact

Users repeat investigation, customers repeat information, and ownership becomes unclear.

Directional evidence

Repetitive administrative work

Meeting preparation, follow-up, reporting, status creation, project planning, resource evaluation, and customer communication are repeatedly recreated.

Impact

Administrative work crowds out strategic judgment and relationship-building.

Validated

Trust gap

Data may be stale, contradictory, incomplete, or disconnected from its source. AI output may not distinguish fact, inference, synthesis, and user-entered information.

Impact

Users re-verify information and are reluctant to act on recommendations.

Validated

Different roles experience different symptoms, but the underlying pattern is the same: too much manual work before confident action.

CSMs

  • Manual meeting preparation
  • Reactive support visibility
  • Repeated account synthesis
  • Customer follow-up burden
  • Difficulty anticipating customer needs

SAMs and Support

  • Fragmented case history
  • Unclear ownership
  • Internal activity not visible to customers
  • Escalation coordination burden

PAs and TSMs

  • Technical context separated from adoption and customer goals
  • Difficulty connecting configuration, usage, cases, and roadmap implications

EMs

  • Project planning outside core systems
  • Duplicate internal and customer reporting
  • Disconnected forecast, budget, timeline, and resource information

RMs

  • Late pipeline visibility
  • Incomplete skills and availability information
  • Manual partner outreach
  • Spreadsheet-based capacity planning

Leadership

  • Manual portfolio synthesis
  • Inconsistent reporting
  • Limited ability to identify emerging risk across accounts

Today’s tool sprawl

CSPNowSupportSurfOrcaTableauTeamsOutlookZoomSharePointExcelPowerPointProduct telemetryCRMCustomer systems

Section 03

Vision and North Star

What CSP becomes, what it always knows, and what it should know right now.

Vision statement

CSP on AINPX is a personalized AI operating environment for the people responsible for customer success.

The platform should understand the customer situation, assemble the right context, involve the appropriate roles, and guide work toward an outcome.

Role-agnostic in structure. Role-aware in presentation. Context-aware in behavior. Permission-aware in execution.

Strategic shift

CSP should stop growing as a collection of role-based workspaces, tabs, dashboards, and individual features. It should become a shared intelligence and orchestration layer that dynamically assembles around the customer moment.

Core hypothesis

If CSP becomes AI-native, users will spend less time searching, stitching, validating, and monitoring information, and more time making decisions, coordinating work, and helping customers achieve outcomes.

When CSP continuously interprets customer context and anticipates emerging needs, CSMs can intervene earlier, prepare less, and spend more time guiding customer outcomes rather than reacting to requests.

North Star

A self-assembling, continuously maintained customer-success environment with explainable memory, trusted decision-point interruptions, and coordinated action across roles.

North StarDesign hypothesis

The platform promise

The platform knows the user, the customer, the current moment, what has changed, what matters, who should be involved, and what decision or action needs to happen next.

What the platform should always know

Portfolio and customer healthPriority accountsOpen risksEscalationsAdoption and usageValue realizationProjects and milestonesCommitmentsOwnersCustomer sentimentUpcoming meetingsService and account relationshipsSource and confidence

What it should know right now

What changedWhy it mattersWhich customer needs attentionWhat is staleWhat is blockedWhat is at riskWhat the customer may ask nextWhat action is recommendedWho owns the actionWhat AI has already preparedWhat requires human approvalWhat evidence supports the recommendation
Instead ofThe platform says
Go find the information.Here is what needs attention.
Monitor every dashboard.You will be interrupted when judgment is needed.
Prepare from a blank page.Review, confirm, and adjust a continuously maintained brief.
Move between role workspaces.Enter one shared environment that adapts to the role and moment.
Ask AI where to find something.AI anticipates what matters and helps move the work forward.

Section 04

Strategic shifts

Six changes to architectural assumptions. Select a shift for the full from-to detail.

Section 05

The future experience model

Detect, Understand, Coordinate, Act, Learn — the loop every customer moment runs through.

Stage 1

Detect

Identify a meaningful customer or operational signal before the user begins searching.

Examples

Adoption declineStale support caseSLA riskNegative sentimentProject milestone riskResource shortageUpcoming executive meetingUnfulfilled customer commitmentUsage pattern indicating unrealized value

Platform

  • Monitor signals
  • Identify meaningful change
  • Rank urgency
  • Consider customer context
  • Avoid unnecessary interruption
Directional evidence

Stage 2

Understand

Assemble the context needed to evaluate the situation.

Context assembled

Customer objectivesAccount historyProduct footprintUsage and adoptionSupport casesProjectsResourcesCommitmentsMeeting historySentimentService tierStakeholdersRoles and ownersRelevant source systems

Platform

  • Explain what changed
  • Connect related information
  • Identify gaps
  • Show source and confidence
  • Distinguish facts from inferences
Directional evidence

Stage 3

Coordinate

Identify who should participate and create a shared path forward.

Platform

  • Determine primary owner
  • Identify supporting roles
  • Recommend next actions
  • Create tasks
  • Route approvals
  • Share relevant context
  • Preserve permission boundaries
  • Avoid duplicate investigation
Design hypothesis

Stage 4

Act

Help each role complete the appropriate work.

Platform

  • Prepare outputs
  • Execute approved workflows
  • Update systems
  • Draft communication
  • Escalate
  • Schedule meetings
  • Replan
  • Assign ownership
  • Track completion

Human

ValidateDecideApproveAdjustCommunicateApply judgment
Design hypothesis

Stage 5

Learn

Update the shared customer and operational context based on the outcome.

Platform

  • Record decisions
  • Capture results
  • Update risk
  • Update account memory
  • Adjust recommendations
  • Improve future prioritization
  • Preserve an audit trail
Design hypothesis

Section 06

The critical customer moment

Primary future-state scenario: anticipate and coordinate a customer risk before the customer asks.

Trigger

A critical customer support case has not received a substantive update. The customer has an executive meeting the following day. Recent sentiment is negative, and the case may be raised during the meeting.

Detect

  • Support Intelligence identifies stale activity.
  • The platform recognizes the upcoming meeting.
  • Sentiment and account importance increase the priority.
  • The system determines that a human decision is needed.

Understand

  • Case priority
  • SLA status
  • Current assignee
  • Last substantive response
  • Internal work notes
  • Customer-visible activity
  • Release and patch information
  • Related customer commitments
  • Upcoming meeting
  • Customer sentiment
  • Previous escalation history
  • Account-team ownership

Coordinate

  • Primary relationship owner: CSM
  • Support coordination: SAM
  • Technical owner: Support Engineer
  • Technical context contributor: PA or TSM
  • Customer-facing approver: CSM
  • Proposed: internal escalation, ownership confirmation, customer communication, meeting talking points, follow-up monitoring

Act

  • SAM receives an escalation task.
  • Support receives the complete case context.
  • The CSM receives a customer-ready brief.
  • Communication capability prepares an external update.
  • The CSM reviews and approves the communication.
  • The customer receives an update before needing to ask.

Learn

  • Case and account context are updated.
  • The outcome is attached to the customer history.
  • The system records which recommendation was accepted.
  • Future risk assessment incorporates the outcome.
  • The account team retains a complete decision trail.

Customer outcome

The customer feels that ServiceNow is informed, coordinated, and proactive.

Operational outcome

Roles act from shared context rather than repeating investigation across tools.

Strategic outcome

CSP demonstrates that AI-native architecture can solve a category of customer-service problems rather than add another support dashboard.

Section 07

Role-agnostic operating model

One shared platform. Different responsibilities. Personalized access and dynamically assembled workflows.

Role-agnostic means

  • One shared platform
  • One connected context model
  • Shared capabilities
  • Shared orchestration
  • Shared interaction patterns
  • No requirement for a separate workspace for every role

Role-aware means

  • Different default information
  • Different recommended actions
  • Different decision authority
  • Different level of detail
  • Different workflow participation
  • Different permissions
  • Different customer-facing responsibilities

Assembly inputs

RoleResponsibilitiesCustomer assignmentAccount tierService modelCustomer momentWorkflow participationData permissionDecision authorityUser preferencesInteraction preferenceRisk and urgency
D

Default

Automatically available because it is central to the user's responsibilities.

C

Contextual

Available when the user becomes involved in a relevant customer, project, case, escalation, or workflow.

R

Requestable

Available through a user-initiated access request.

A

Approval required

Requires authorization before data can be viewed or an action can be completed.

X

Restricted

Unavailable because it falls outside responsibility, policy, privacy, or security boundaries.

The platform should reveal capabilities when they become relevant without exposing every tool, agent, action, or dataset to every user.

CSM lens

Customer prioritiesAdoptionValueRelationship riskSupport implicationsRecommended customer action

SAM lens

Support healthSLA riskCase stalenessEscalation coordinationSupport trends

PA lens

Technical healthProduct configurationArchitectureTechnical debtRoadmap guidance

EM lens

Delivery healthMilestonesResourcesHoursBudgetCustomer status

RM lens

PipelineCapacitySkillsAvailabilityStaffing riskPartner options

Section 08

Intelligence architecture

Seven layers from source systems to measurable outcomes. Select a layer for inputs, outputs, dependencies, and trust requirements.

Section 09

The Three Awareness types

System, workflow, and role awareness together make anticipation possible.

System Awareness

What is happening?

Customer stateAccount healthProduct usageAdoptionCasesProjectsResourcesCommitmentsRelationshipsSentimentBusiness outcomes

Example

A P1 case has not received a substantive update in five days.

Validated

Workflow Awareness

What is affected, and what should happen next?

DependenciesStalled workUpcoming milestonesOwnershipApprovalsRisksEscalation pathsNext-best actionsLikely customer response

Example

The case may become an executive escalation because the customer has a meeting tomorrow and sentiment is declining.

Directional evidence

Role Awareness

What does this user need to know or do now?

ResponsibilitiesAssigned customersDecision authorityRequired level of detailCustomer-facing ownershipAccess rightsApproval requirementsPersonal priorities

Example

The CSM needs the customer impact and communication plan, while Support needs technical history and the SAM needs escalation status.

Directional evidence

Anticipation requires all three forms of awareness. The system must understand the customer state, the workflow implications, and the responsibilities of the person receiving the information.

Section 10

Interaction modes

Hybrid is the primary future-state interaction model because it combines proactive intelligence, structured evidence, workflow controls, and conversation.

Conversational

Best for

ExplorationFollow-up questionsOpen-ended inquiryScenario comparisonDraftingExplanationFinding supporting context

Example

“What should I know before tomorrow's meeting?”

Conversation alone can hide structure, state, completeness, and available actions.

Validated

UI-driven

Best for

Structured recordsDeterministic tasksComparisonHigh-volume monitoringApprovalEditingFinancial and operational workAuditable workflows

Example

A prioritized portfolio list with support risk, SLA, owner, and available actions.

Static UI can reproduce current navigation and dashboard problems if it is not dynamically assembled.

Validated

Hybrid

Dominant

Best for

Proactive briefsDynamic workspacesRisk investigationMeeting preparationGuided workflowsReview and approvalCross-role coordinationAction-oriented recommendations

Example

The platform assembles a support-risk brief with evidence, roles, actions, and a conversational area for follow-up questions.

Directional evidence

Mode selection depends on

Task typeComplexityRiskNeed for structureNeed for explanationRequired actionUser preferencePermissionAuditability

Section 11

Agents and orchestration

Capabilities operate behind one experience. Select an agent for signals, outputs, approvals, and visibility.

  • Agents may operate behind a unified Otto experience.
  • Users should not need to know which agent owns every step.
  • Agent identity should be visible when it supports transparency, control, troubleshooting, or trust.
  • The platform should orchestrate agents around user goals and customer moments.
  • Agents should not compete for attention or create disconnected destinations.

Orchestration in the primary scenario

  1. 1Customer IntelligenceUnderstands account importance, sentiment, and meeting context.
  2. 2Support IntelligenceDetects stale case activity and SLA risk.
  3. 3Risk and OpportunityDetermines likely customer impact.
  4. 4InterlockIdentifies appropriate internal roles.
  5. 5CommunicationPrepares the external update.
  6. 6Reporting and InsightsUpdates the customer and portfolio history.

Section 12

Current tools to future capabilities

Source systems remain. What they are used for changes.

ToolUsed today forFuture capabilityAgents
NowSupport
CasesActivitySLAAssigneeEscalation detailsWork notes
Anticipate and coordinate customer support concerns.
Support Intelligence
Surf
SalesOpportunityContractCustomer commitmentsPipeline
Understand upcoming customer and delivery demand.
Customer IntelligenceRisk and Opportunity
Orca
ProjectsResourcesDeliveryHoursBudgetAllocation
Coordinate delivery, project health, staffing, and resource risk.
Project DeliveryResource Recommendation
Teams and Outlook
Informal decisionsRelationship contextCustomer communicationInternal coordinationSentiment
Preserve account memory and coordinate work.
Meeting PreparationCommunicationInterlock
Zoom
MeetingsTranscriptsParticipant context
Maintain meeting history, decisions, commitments, and follow-up.
Meeting Preparation
Tableau
ReportingPortfolio viewsProject healthOperational trends
Surface prioritized insights and current status.
Reporting and Insights
Excel
ForecastingPlanningResource analysisProject trackingWorkarounds
Provide dynamic planning, scenario modeling, and structured comparison.
Project DeliveryResource Recommendation
PowerPoint
Customer statusExecutive reviewQBR and EBR materials
Generate customer-ready narratives and adaptable outputs.
Reporting and InsightsCommunication
Product telemetry
UsageAdoptionConfigurationProduct health
Identify risk, opportunity, and next-best action.
Adoption and Value
Current sourcesShared contextAwarenessOrchestrationCapabilities and agentsPersonalized experienceCoordinated actionCustomer outcome

Section 13

Time horizons

Now, Next, and North Star are separated so the vision is not read as an immediate roadmap commitment.

Now

Now

Foundation

  • Improve visibility
  • Unify critical information
  • Introduce AI-generated summaries
  • Reduce tool switching
  • Surface source and freshness
  • Improve current workflows
  • Support existing CSP structures

Example capabilities

Account summariesSupport summariesMeeting summariesSource-aware insightsImproved portfolio prioritizationUnified case contextBetter filtering and search

Interaction

UI-driven with embedded AI

Why this matters

Useful, achievable, and compatible with current workflows.

Next

Next

Assisted workflows

  • Connect signals
  • Identify risks
  • Recommend actions
  • Prepare outputs
  • Coordinate work across roles
  • Route approvals
  • Preserve context through handoffs
  • Maintain customer memory

Example capabilities

Proactive risk briefsMeeting lifecycle automationSupport escalation coordinationAdoption interventionsProject-risk recommendationsResource recommendationsCross-role action plans

Interaction

Hybrid

Why this matters

Demonstrates measurable operational value and builds trust.

North Star

North Star

Assembled experiences

  • Dynamically construct the environment around the customer moment
  • Use ambient watching
  • Anticipate emerging needs
  • Assemble relevant roles and capabilities
  • Execute approved multi-step workflows
  • Maintain explainable memory
  • Personalize by role, context, and preference
  • Continuously learn from outcomes

Interaction

Adaptive hybrid, conversational, and UI-driven experiences

Why this matters

Shows the complete AI-native vision without presenting it as an immediate implementation commitment.

Section 14

MVP recommendation and validation

Prove the architecture through one high-value moment.

MVP statement

Prove that CSP can tell a CSM what needs attention today, why it matters, where the information came from, who should be involved, and what action should happen next.

Scenario: Stale critical support case before a customer meeting.

Detection

  • Identify meaningful stale-case activity
  • Connect the case to the customer and upcoming meeting

Context

  • Priority
  • SLA risk
  • Current owner
  • Last substantive response
  • Customer-visible status
  • Customer sentiment
  • Meeting context
  • Source and freshness

Recommendation

  • Explain likely customer impact
  • Recommend escalation
  • Recommend communication
  • Identify roles

Coordination

  • Route internal task
  • Share context
  • Assign ownership

Action

  • Prepare customer communication
  • Require CSM approval
  • Update account context

Trust

  • Show sources
  • Show confidence
  • Distinguish factual and inferred information
  • Allow correction
  • Record approval

Explicitly out of scope

  • Fully autonomous external communication
  • Universal coverage of every CSP role
  • Complete replacement of source systems
  • Unrestricted cross-role access
  • Autonomous handling of sensitive escalations
  • Complete predictive accuracy
  • Final agent naming and visibility model

Validation focus

Desirability

  • Does the signal help users intervene earlier?
  • Is the recommendation relevant?
  • Does the assembled context reduce preparation?
  • Does the flow match actual role responsibility?

Trust

  • Do users understand why the signal appeared?
  • Are sources and freshness sufficient?
  • Do users know what is fact versus inference?
  • Do users feel in control?

Usability

  • Can users understand the priority?
  • Can users inspect evidence?
  • Can users complete the recommended action?
  • Can users distinguish internal and customer-facing content?

Feasibility

  • Are required sources available?
  • Is the data sufficiently fresh?
  • Can ownership be resolved?
  • Can actions be executed through connected systems?
  • Are permissions enforceable?

Value

  • Does the flow reduce manual work?
  • Does it shorten time to intervention?
  • Does it improve cross-role coordination?
  • Does it reduce customer effort?

Section 15

Value and outcomes

What changes for customers, roles, leadership, and the platform itself.

For customers

  • ServiceNow appears informed and coordinated
  • Customers repeat less information
  • Risks are addressed earlier
  • Communication becomes more proactive
  • Support and success feel continuous
  • Customer effort decreases
  • Value realization becomes clearer

For CSMs

  • Less context gathering
  • Faster preparation
  • Earlier risk awareness
  • Better account prioritization
  • Stronger customer conversations
  • More time for strategic guidance
  • Reduced administrative burden

For specialized roles

  • Better-quality handoffs
  • More relevant context
  • Clearer ownership
  • Fewer ad hoc status requests
  • Reduced duplicate investigation
  • Improved collaboration

For leadership

  • More consistent service
  • Better portfolio visibility
  • Greater capacity
  • Earlier risk detection
  • More measurable outcomes
  • Better alignment across service tiers

For product and engineering

  • Fewer bespoke role workspaces
  • Reusable architecture
  • Shared capability model
  • Lower long-term UI maintenance
  • Clearer assembly rules
  • Scalable agent orchestration

For design and research

  • Context models
  • Assembly logic
  • Interaction rules
  • Agent behavior
  • Trust patterns
  • Role lenses
  • Workflow orchestration
  • Evidence-backed future states

Section 16

Success measures

Directional measures grouped by category. Status shows how ready each measure is today.

User efficiency

  • Time spent gathering customer contextNowBaseline needed
  • Time spent preparing for meetingsNowBaseline needed
  • Time spent creating status updatesNextProposed
  • Systems opened per workflowNowAvailable
  • Reduction in duplicate entryNextProposed
  • Reduction in manual follow-upNextProposed

Proactive service

  • Risks identified before customer escalationNextProposed
  • Time between signal detection and human actionNextIn development
  • Customer updates sent before customer requestNextProposed
  • Preventable escalations avoidedNorth StarProposed
  • Proactive versus reactive intervention ratioNorth StarProposed

Workflow coordination

  • Time to establish ownershipNextBaseline needed
  • Handoff completion rateNextProposed
  • Context completeness at handoffNextProposed
  • Duplicate investigation rateNextBaseline needed
  • Cross-role task completionNorth StarProposed
  • Approval-cycle timeNextAvailable

Trust

  • Recommendation acceptanceNowIn development
  • Recommendation correctionNowIn development
  • Source-inspection rateNowProposed
  • Confidence understandingNextProposed
  • User trust scoreNextBaseline needed
  • AI output requiring re-verificationNowProposed

Customer outcomes

  • Customer Effort ScoreNextAvailable
  • Customer satisfactionNowAvailable
  • Support satisfactionNowAvailable
  • Adoption improvementNextAvailable
  • Time to valueNorth StarBaseline needed
  • Risk resolutionNextProposed
  • Value realizationNorth StarProposed
  • Retention or expansion influenceNorth StarProposed

Platform outcomes

  • Role-specific surfaces retired or avoidedNorth StarProposed
  • Reuse of shared capabilitiesNextProposed
  • Agent orchestration reuseNorth StarProposed
  • Reduction in navigation complexityNextBaseline needed
  • Workflows supported through shared contextNorth StarProposed
  • Assembled versus fixed experiencesNorth StarProposed

Section 17

Experience principles

Ten design rules, each with what good looks like and the anti-pattern it replaces.

1

Organize around customer moments, not product navigation.

The organizing unit of the experience is a situation, not a screen.

Why: Navigation growth is the root cause of the current complexity.

What good looks like

An escalation assembles case, meeting, sentiment, and owners in one place.

Anti-pattern

Adding an escalation tab to every role workspace.

North StarDesign hypothesis
2

Anticipate before asking, but interrupt only when action or judgment is needed.

Detection is continuous; interruption is rare and justified.

Why: Alert volume destroys trust faster than missing signals.

What good looks like

One high-confidence interruption before an executive meeting.

Anti-pattern

A daily digest of every changed field.

NextDirectional evidence
3

Assemble context rather than sending users to gather it.

The platform does the stitching work before the user arrives.

Why: Context gathering is the largest reported time cost.

What good looks like

A support-risk brief already contains SLA, owner, and last response.

Anti-pattern

A link list to five source systems.

NowValidated
4

Recommend action, not only information.

Every insight names a next step and an owner.

Why: Users value action-oriented output over summaries.

What good looks like

“Escalate to SAM and send the customer an update” with both actions available.

Anti-pattern

A summary paragraph with no available action.

NextValidated
5

Maintain one shared context while adapting presentation by role.

One context model, many lenses.

Why: Duplicated role surfaces are the maintenance problem being solved.

What good looks like

CSM and SAM see the same case through different defaults.

Anti-pattern

A separate SAM workspace with its own data model.

NextDesign hypothesis
6

Keep humans accountable for sensitive decisions and customer commitments.

AI prepares; people approve anything the customer will see.

Why: Accountability is a precondition for adoption.

What good looks like

A drafted customer update waits for CSM approval.

Anti-pattern

Autonomous external communication.

NowValidated
7

Make facts, inferences, sources, freshness, and confidence visible.

Every statement can be inspected back to its origin.

Why: Users re-verify anything they cannot trace.

What good looks like

An inferred risk is labelled and links to the signals behind it.

Anti-pattern

A confident sentence with no source.

NowValidated
8

Preserve context across roles, systems, and time.

Handoffs carry the full picture, not a link.

Why: Lost context is the main cost of cross-role work.

What good looks like

Support receives the assembled case package with the task.

Anti-pattern

A Teams message saying “please look at this case”.

NextDirectional evidence
9

Use conversation as one interaction mode, not the entire experience.

Conversation supports structured surfaces; it does not replace them.

Why: Blank conversational interfaces were insufficient in research.

What good looks like

A brief with a follow-up conversation area.

Anti-pattern

A chat box as the primary entry point.

NowValidated
10

Solve reusable categories of problems instead of adding isolated features.

Each solution should generalize to other moments and roles.

Why: Feature-by-feature growth created the current cost curve.

What good looks like

The stale-case pattern reused for stale commitments and stale milestones.

Anti-pattern

A single-purpose stale-case dashboard.

North StarDesign hypothesis

Section 18

Risks and assumptions

What could go wrong, how it is mitigated, and which assumptions still need validation.

The vision becomes another CSM workspace.

Mitigation

Anchor the story with the CSM but demonstrate reusable architecture, shared context, and cross-role participation.

Open question

Role-agnostic is interpreted as identical access.

Mitigation

Use role-aware defaults, contextual access, requestable capabilities, approvals, and restrictions.

Directional evidence

AI produces too many alerts.

Mitigation

Focus on decision-point interruption, relevance, confidence, and user-controlled thresholds.

Validated

Users do not trust the data.

Mitigation

Show source, freshness, confidence, verification status, and correction pathways.

Validated

Conversational AI becomes the primary design metaphor.

Mitigation

Use hybrid experiences with structured context, evidence, controls, and actions.

Directional evidence

Agents become disconnected destinations.

Mitigation

Orchestrate agents behind customer moments and unified user goals.

Design hypothesis

Cross-role coordination ignores actual responsibility.

Mitigation

Validate ownership, authority, handoff, and service-tier differences with representative users.

Open question

The North Star is interpreted as an immediate roadmap commitment.

Mitigation

Clearly separate Now, Next, and North Star horizons.

Directional evidence

Personalization becomes unpredictable.

Mitigation

Explain why content is shown and allow users to inspect, adjust, pin, dismiss, or request capabilities.

Design hypothesis

The model depends on unavailable or poor-quality data.

Mitigation

Identify source dependencies, quality gaps, confidence thresholds, and fallback states.

Open question
AssumptionEvidenceConfidenceStatusOwnerValidation
CSMs benefit from proactive customer-risk intelligence.Support Intelligence research: stale case detection rated highly valuable.ValidatedSupportedResearchConcept test with guided and self-serve CSMsBy Before MVP build
Hybrid interaction will be more useful than conversation alone.CSP AI Workflows research: blank conversational interfaces were insufficient.Directional evidenceSupportedDesignComparative prototype studyBy Before MVP build
Shared context can support multiple roles without duplicating the interface.Design hypothesis derived from the role lens model.Design hypothesisOpenProduct and EngineeringArchitecture spike plus cross-role reviewBy Next planning cycle
Users will accept dynamic assembly when they understand why information is shown.Directional evidence on explanation and trust.Directional evidenceOpenResearchUsability study with explanation variantsBy Before North Star commitment
Agents can coordinate across systems while preserving human approval.Unproven at scale.Design hypothesisOpenEngineeringTechnical feasibility spikeBy Before Next horizon
Current tools can remain systems of record while CSP becomes the operating environment.Integration dependencies not fully assessed.Open questionOpenPlatformSource dependency auditBy Before MVP build
Customer moments are a more scalable organizing principle than persona workspaces.Vision strategy position; conflicting internal views exist.Conflicting evidenceOpenProductStakeholder alignment workshopBy This quarter
Role and permission models can support contextual access.Access state model is proposed, not implemented.Design hypothesisOpenPlatform and SecurityPermission model reviewBy Before Next horizon

Section 19

Strategic decisions to make

Open questions this vision is intended to force, grouped by decision area.

Vision scope

  • Which customer moment carries the highest strategic value?

    ProposedProduct
  • Should risk and escalation remain the primary scenario?

    ProposedProduct
  • How broadly should the first concept represent non-CSM roles?

    OpenDesign

Product direction

  • Which capabilities are foundational?

    OpenProduct
  • Which capabilities are near-term, emerging, or aspirational?

    OpenProduct
  • Which existing surfaces should evolve, remain, or eventually retire?

    OpenProduct
  • How does the vision align with Project Venus, CRMI, AINPX, and Horizon 2.0?

    OpenStrategy

Experience direction

  • Should users see individual agents or one unified Otto experience?

    OpenDesign
  • What should be proactively surfaced?

    ProposedDesign
  • What should remain available on demand?

    ProposedDesign
  • When should the platform use conversational, hybrid, or UI-driven modes?

    AlignedDesign
  • How much personalization should the MVP include?

    OpenProduct

Data and trust

  • Which sources are authoritative?

    OpenPlatform
  • What freshness thresholds are acceptable?

    OpenPlatform
  • How should conflicting data be presented?

    OpenDesign
  • What requires user confirmation?

    ProposedDesign
  • How should inference and prediction be labeled?

    ProposedDesign

Role and governance

  • Which roles require default access?

    ProposedProduct
  • What becomes contextual or requestable?

    OpenProduct
  • Who approves temporary access?

    OpenSecurity
  • Which actions require explicit human approval?

    AlignedProduct
  • Who owns the customer-facing communication?

    AlignedService

Validation

  • Which PM, design, engineering, research, and user representatives must review the concept?

    OpenProgram
  • What evidence is required before the strategy becomes a shared direction?

    OpenResearch
  • Which assumptions must be resolved before prototyping?

    OpenResearch
  • Which assumptions can be tested through the prototype?

    ProposedResearch

Section 20

Research traceability

Where the strategy is grounded in evidence, and where it remains a hypothesis.

Future of Service at ServiceNow

Directional evidence
  • Service is both a strategic risk and differentiating opportunity.
  • The future model should focus on relationships, moments, and outcomes.
  • AI should enhance people.
  • Service should be evaluated end to end.
  • Trust and continuity must persist across surfaces.
  • The organization needs a bridge from current case-centric service to a future service model.

CSP AI Workflows Research

Validated
  • Context is fragmented and often inaccurate.
  • Users spend significant time gathering information before customer interactions.
  • Administrative work prevents proactive customer engagement.
  • Users want AI to surface what they need before they ask.
  • Users value action-oriented outputs.
  • Customer Touchpoint Automation received strong interest.
  • Blank conversational interfaces are insufficient.

Support Intelligence Research

Validated
  • CSMs lack a consolidated view of support cases.
  • Guided CSMs experience greater support-monitoring burden.
  • Stale case detection is highly valuable.
  • Users want priority, owner, last activity, SLA, and case summary.
  • Users prefer a prepared snapshot with drill-down and follow-up questions.
  • Accuracy is critical for trust.

Expert Services Research

Directional evidence
  • RMs and EMs work across fragmented tools.
  • RMs need early pipeline awareness.
  • Skills and availability data are incomplete.
  • Partner sourcing is manual.
  • EMs use external spreadsheets for hours, budgets, and project plans.
  • Internal and customer-facing reporting is duplicated.
  • Cross-role context is required for successful delivery.

CSP on AINPX Vision Strategy

Design hypothesis
  • Use a moment-led approach.
  • Anchor with the CSM but reveal cross-role orchestration.
  • Use the Three Awareness model.
  • Demonstrate one deep scenario.
  • Separate Now, Next, and North Star.
  • Connect the future concept to reusable architecture.
Strategy elementSupporting research
From feature-by-feature growth to architecture-led intelligence
CSP on AINPX Vision Strategy
From navigated views to assembled experiences
CSP on AINPX Vision StrategyCSP AI Workflows Research
From reactive dashboards to ambient watching
Support Intelligence Research
From role-specific tools to a shared context graph
Expert Services ResearchCSP on AINPX Vision Strategy
From tribal knowledge to durable, explainable memory
CSP AI Workflows ResearchSupport Intelligence Research
From periodic preparation to a continuously ready state
CSP AI Workflows Research
Fixed experience growth
CSP AI Workflows Research
Fragmented context
CSP AI Workflows ResearchExpert Services Research
Reactive monitoring
Support Intelligence Research
Incomplete handoffs
Expert Services Research
Repetitive administrative work
CSP AI Workflows ResearchExpert Services Research
Trust gap
Support Intelligence Research

Current state to future state

Compare how the operating model changes across nine dimensions.

Shared contextDynamic assemblyAmbient watchingExplainable memoryGuided actionCross-role orchestrationPersonalized accessContinuously ready context