Zava Innovation Hub

Release candidate: This offline, mock-data sample demonstrates how Microsoft 365 Copilot can turn natural-language intent into focused, interactive business UX. Inline components complete immediate work; full screen preserves that context while adding application-scale comparison and governance. The implementation follows agentic-creation-rules.md; current validation status and external prerequisites are tracked in todo.md.

Screenshots

Enterprise Insights	Global participation

Idea submission	Review queue

Business story

An employee has an idea. The organization helps shape it, makes a governed investment decision, learns from the pilot, realizes value, and celebrates the people who made it happen.

Most innovation portals make employees navigate forms, lists, dashboards, and approval sites. This sample shows a different model:

1. A person describes a need naturally in Copilot.
2. Copilot selects one bounded experience and initializes it from the conversation.
3. The person works in a purpose-built inline form, review, decision, or visualization.
4. Deterministic UX owns calculations, validation, state changes, and confirmation.
5. Full screen continues from the exact idea, cohort, filter, draft, or funding scenario when broader context is useful.

The domain is innovation management, but the reusable message is dynamic UX based on user intent. The same agent can present a submission workflow, investment model, review queue, interactive portfolio visual, or recognition moment without sending the user to another application.

Market-informed product position

This is a directional design benchmark, not a procurement comparison or claim of feature parity. Public product pages reviewed on 2026-08-27 show a consistent enterprise innovation baseline:

• Brightidea emphasizes challenges, global employee collaboration, idea pipelines, stage gates, evaluation, social collaboration, analytics, and ROI tracking.
• HYPE positions innovation as a connection between strategy, ideation, ecosystems/scouting, execution, analytics, and measurable impact across several innovation roles.
• Planview IdeaPlace describes challenge templates, idea capture, voting/collaboration, consolidation, evaluation workflows, portfolio/project handoff, engagement, and executive reporting.
• Wazoku describes challenge-driven workflows, communities and co-creation, analytics, AI-assisted duplication/content discovery, portfolio management, and integration with enterprise tools.
• Qmarkets frames an end-to-end ecosystem spanning ideas, continuous improvement, foresight, technology scouting, portfolios, configurable processes, and impact defined from the beginning.

Reference pages: Brightidea, HYPE Innovation, Planview IdeaPlace, Wazoku Platform, and Qmarkets.

The market validates the need for a complete operating system, but this sample should not imitate a traditional portal screen for screen. Its differentiator is the agentic interaction model:

Traditional platform pattern	Copilot Components modernization
People learn navigation, forms, taxonomies, and process location.	People state the job naturally; Copilot selects one bounded component and initializes the right entity, scope, or draft.
Intake begins when someone finds the correct campaign or form.	A conversational signal can become a challenge, idea draft, duplicate check, or evidence request in the current flow of work.
Dashboards are destinations users must visit and configure.	A decision-specific chart appears inline for the current intent; full screen adds coordinated analysis without losing context.
Workflow automation moves records but often separates evidence from decisions.	The review component assembles exact evidence and consequences, while the human owns the deterministic decision and confirmation.
Collaboration depends on returning to the innovation portal.	Follow-ups, feedback, praise, and next actions can surface where people already work, with future Teams/Graph connectors behind stable UI contracts.
AI is commonly an assistant inside the platform.	AI performs intent resolution and bounded handoff; components own trusted records, calculations, controls, safeguards, and receipts.

The sample is therefore not a replacement claim for mature innovation suites. It demonstrates how their end-to-end operating model can become lower-friction, role-aware, and conversation-native while remaining governable and integration-ready.

End-to-end operating model

The design covers the lifecycle from strategic demand to measured impact, not merely an idea inbox:

Lifecycle phase	Business job	Agentic experience	Current owner
1. Direction	Translate strategy, customer pain, or operational need into a measurable challenge.	Copilot shapes an outcome, audience, constraints, evaluation criteria, timeline, and challenge brief.	LaunchInnovationChallenge
2. Signals and opportunity	Bring observations, trends, prior ideas, and external signals into the problem.	Relevant internal evidence and similar ideas appear in context; external scouting is a future service-backed extension.	Challenge and submission dynamic evidence
3. Ideation and participation	Capture and improve contributions from employees and collaborators.	Prompt-prefilled idea canvas, duplicate evidence, scoring, feedback, supporters, and global participation views.	SubmitInnovationIdea, GetMyInnovation, ExploreGlobalInnovation
4. Triage and shaping	Separate promising concepts from noise and strengthen their case.	Ranked queue, cohort comparison, evidence gaps, gate decision, and live business-case modeling.	GetInnovationReviewQueue, ReviewIdeaGate, BuildIdeaBusinessCase
5. Investment	Allocate scarce capital against strategy and portfolio constraints.	Funding scenario and confirmation beside budget, balance, value, and milestone consequences.	ReviewInnovationFunding, TrackInnovationBudget
6. Experiment and learn	Turn assumptions into a measurable pilot and make evidence-based go/no-go decisions.	Hypothesis, experiment design, success thresholds, milestones, observations, learning, and guarded recommendation.	ManageInnovationExperiment
7. Scale and execute	Move validated concepts toward rollout while retaining strategic and financial traceability.	Stage progression, milestone evidence, ownership, and portfolio context continue from the same idea graph.	Experiment, portfolio, and value routes
8. Realize value	Compare promised outcomes with measured operational or financial impact.	Projected-versus-actual value, confidence, variance, at-risk outcomes, and accountable owners.	TrackInnovationValue
9. Learn and recognize	Reuse knowledge and celebrate specific contributions that produced outcomes.	Grounded executive brief, reusable evidence, and a reviewed recognition artifact tied to realized value.	GenerateInnovationBrief, CelebrateInnovationImpact
10. Govern and improve	Improve reach, velocity, balance, inclusion, and return across the innovation system.	Growth, geography, portfolio health, and leadership views reveal where the operating model needs action.	GetInnovationGrowth, ExploreInnovationPortfolio, GetInnovationPortfolioHealth

The current sample supports an employee/corporate innovation program deeply. Technology scouting, startup/partner ecosystems, patents, customer crowds, configurable methodology administration, and durable project-system execution are documented extension points. They should reuse the same scenario, workflow, and view-model contracts rather than become decorative placeholder tools.

Showcase objectives

Objective	Keynote proof
Make inline UX the hero	Distinct prompts resolve to visibly different, interactive components rather than generic generated cards.
Demonstrate meaningful adaptation	Prompt parameters and controls materially change records, calculations, chart geometry, selections, or workflow stages.
Complete work in conversation	Idea submission, gate review, funding review, and recognition reach explicit review and mock confirmation inside Copilot.
Show exact continuation	View in full screen opens the owning lens and focused route with current context intact.
Tell one governed story	Connected data supports intake, triage, business case, funding, pilot, value realization, and recognition.
Deliver premium visual quality	A living funnel, portfolio visuals, accountable people, deliberate motion, and editorial composition create a keynote-ready experience.

Success is not measured by component count. Each routed component must earn its place through at least two of these qualities: frequent or time-sensitive value, meaningful prompt variation, useful interaction or review, distinctive visualization, and natural full-screen continuation.

Audiences and lenses

Lens	Demo persona	Core question
My Innovation	Megan Bowen, Senior Program Manager and innovator	What should I submit or strengthen, where are my ideas, and how is my contribution recognized?
Programs & Pilots	Johanna Lorenz, Product Director and innovation program lead	Which challenges should we run, where is participation weak, and what are pilots teaching us?
Reviews & Gates	Diego Siciliani, Engineering Manager and council reviewer	Which ideas deserve the next gate, and what evidence supports the decision?
Investment	Miriam Graham, Finance Business Partner	Where is innovation funding committed, forecast, constrained, and available for the next decision?
Enterprise Insights	Joni Sherman, VP of Digital Strategy and portfolio sponsor	Is innovation growing across the company, globally inclusive, balanced, and realizing value?

The five primary personas and six supporting specialists reuse the bundled reference portraits below. Use a portrait where a person owns an idea, challenge, review, investment, experiment, or outcome; otherwise use Fluent Persona initials to keep analytical views focused.

Bundled persona assets

All files are copied byte-for-byte from the proven zava-project-tracker/assets/faces reference set and stored locally under assets/faces; no runtime profile-photo request is required.

Person	Reference role / use in this sample	Local portrait
Megan Bowen	Senior Program Manager; primary innovator and idea owner	assets/faces/Megan-Bowen.jpeg
Johanna Lorenz	Product Director; primary challenge and pilot-program lead	assets/faces/Johanna-Lorenz.jpeg
Diego Siciliani	Engineering Manager; primary gate reviewer and council member	assets/faces/Diego-Siciliani.jpeg
Miriam Graham	Finance Business Partner; primary investment and budget owner	assets/faces/Miriam-Graham.jpeg
Joni Sherman	VP, Digital Strategy; primary executive and portfolio sponsor	assets/faces/Joni-Sherman.jpeg
Pradeep Gupta	AI Platform Lead; technical evaluator and experiment specialist	assets/faces/Pradeep-Gupta.jpeg
Nestor Wilke	Responsible AI Lead; risk, governance, and criteria specialist	assets/faces/Nestor-Wilke.jpeg
Lee Gu	UX and Adoption Lead; participation and adoption specialist	assets/faces/Lee-Gu.jpeg
Patti Fernandez	Change Lead; challenge communications and recognition partner	assets/faces/Patti-Fernandez.jpeg
Isaiah Langer	Data Engineering Lead; measurement and experiment-data owner	assets/faces/Isaiah-Langer.jpeg
Grady Archie	Procurement Lead; external-partner and supplier-evidence specialist	assets/faces/Grady-Archie.jpeg

The implementation media catalog must use these exact person names and case-sensitive filename keys. Initials remain the loading/missing-media fallback. Source rights and redistribution evidence must be recorded in the asset provenance manifest before public release.

End-to-end journey

stateDiagram-v2
    [*] --> Draft
    Draft --> Submitted: Review and confirm
    Submitted --> Screening
    Screening --> BusinessCase: Gate 1 advance
    Screening --> NeedsWork: Request changes
    Screening --> Rejected
    NeedsWork --> Screening: Resubmit
    BusinessCase --> FundingReview: Gate 2 advance
    BusinessCase --> Parked
    FundingReview --> Funded
    FundingReview --> PartiallyFunded
    FundingReview --> Deferred
    FundingReview --> Rejected
    Funded --> Pilot
    PartiallyFunded --> Pilot
    Pilot --> Scaling: Go
    Pilot --> Sunset: No-go
    Scaling --> ValueRealized
    ValueRealized --> Celebrated
    Celebrated --> [*]

The keynote follows one connected idea, Smart Onboarding Journey:

1. Megan submits it; duplicate evidence and a five-axis score appear inside the submission flow.
2. Diego compares it with its cohort and advances it after reviewing gate evidence.
3. Megan shapes the business case; investment, benefit, payback, and sensitivity update live.
4. Miriam reviews funding; changing the amount redraws portfolio mix and expected value before an explicit decision is confirmed.
5. Johanna runs the pilot experiment; evidence changes the recommendation from Go to Pivot.
6. Joni's command center updates from the session-only receipts.
7. A realized outcome becomes a beautiful recognition card for Megan and her collaborators.

Target inline catalog

The target is 16 operational Copilot Components plus one capability explorer. The earlier proposal for 30 separate components is superseded: duplicate detection, score summaries, ROI/NPV, sponsor detail, portfolio balance checks, lifecycle progress, voting, and related charts are dynamic modules inside the owning experience, not independent conversational tools. Challenge framing, experimentation, growth, geography, budget, and leadership health remain independent because they answer different roles, decisions, time horizons, and prompts.

Model	Count	Contract
Information / interactive analysis	8	Answer immediately, then allow material filtering, selection, comparison, or chart-mode changes.
Review / decision	3	Queue or selected record -> evidence -> decision draft -> confirmation -> session receipt.
Submit / create	5	Prompt-prefilled draft -> validation -> review -> confirmation -> session receipt.
Education / discovery	1	Search and filter every operational scenario, copy a prompt, and preview safely.

#	Final component / tool	Distinct inline value and dynamic behavior	Full-screen continuation
1	SubmitInnovationIdea	Guided idea canvas for problem, audience, outcome, theme, collaborators, and evidence. Editing updates similarity, five-axis score, readiness, and required evidence. Review and explicit mock submission are required.	my-innovation/new-idea, preserving draft and step.
2	GetMyInnovation	Personal portfolio showing each idea on the living funnel with next action, feedback, supporters, milestones, and recognition. Filters change the set; selection synchronizes detail.	my-innovation/overview, preserving idea and filters.
3	BuildIdeaBusinessCase	Editable investment model with benefit drivers, 36-month cash flow, payback, ROI/NPV, uncertainty, and milestones. Assumptions recalculate the model live.	my-innovation/business-case, preserving idea and scenario.
4	CelebrateInnovationImpact	Evidence-grounded recognition composer with recipients, collaborators, outcome, badge, praise theme, audience, and polished share preview. Review precedes a mock share receipt.	my-innovation/recognition, preserving composition.
5	GetInnovationReviewQueue	Ranked reviewer inbox with owner, age, gate, evidence completeness, score, conflicts, and due state. Filters alter ranking and counts; selection opens concise evidence.	reviews-gates/review-queue, preserving filters and selection.
6	ReviewIdeaGate	One-at-a-time submission review with phase criteria, evidence completeness, strengths, gaps, owner, score, and consequence. Approve advances, Send back requests more evaluation, and Decline closes the submission; every outcome requires review, confirmation, and a session receipt.	reviews-gates/gate-review, preserving phase, submission, rationale, and decision draft.
7	ReviewInnovationFunding	Signature decision: business economics beside before/after portfolio balance and funding controls. Proposed amounts update mix, budget, value, and milestone scope before confirmation.	investment/funding-committee, preserving request and amount scenario.
8	ExploreInnovationPortfolio	Executive analytical canvas that switches among living funnel, bubbles, horizon balance, theme map, and funding bridge based on intent or direct control. Filters rebuild the analytical model.	enterprise-insights/command-center, preserving filters, mode, and selected mark.
9	TrackInnovationValue	Funded-pilot milestones, projected-vs-actual value, variance, risk to value, and accountable owner. Filters and selection coordinate timeline and value evidence.	enterprise-insights/value-realization, preserving cohort, metric, and pilot.
10	GenerateInnovationBrief	Grounded executive brief builder with editable audience, period, emphasis, selected evidence, and visual cover statistic. Copy, mock Send, and Export stop at review.	enterprise-insights/executive-brief, preserving narrative settings.
11	GetInnovationGrowth	Innovation Operations trend view showing submissions, active contributors, gate throughput, conversion, and realized ideas over time. A layered line/area chart and cohort bands distinguish healthy growth from intake volume alone.	programs-pilots/growth, preserving metric, period, cohort, and selected point.
12	ExploreGlobalInnovation	Global participation view using a locally packaged world map, proportional submission marks, regional conversion bars, theme mix, and inclusion gaps. Region/metric selection changes both geography and evidence.	programs-pilots/geography, preserving metric, region, theme, and selected location.
13	TrackInnovationBudget	Finance-owned budget control with allocation, committed, approved, spent, forecast, and remaining funding. A bridge chart, burn/forecast line, and horizon allocation react to period and scenario.	investment/budget, preserving period, horizon, scenario, and selected variance.
14	GetInnovationPortfolioHealth	Leadership scorecard combining funnel velocity, portfolio balance, strategic alignment, concentration risk, expected value, and realized value. Selecting a health dimension reveals its trend and accountable exceptions.	enterprise-insights/leadership-health, preserving period, dimension, and selected exception.
15	LaunchInnovationChallenge	Innovation-program workflow that turns a strategic priority or business problem into a measurable challenge brief: outcome, audience, constraints, evidence, criteria, timeline, sponsor, and participation plan.	programs-pilots/challenge-studio, preserving challenge draft, audience, criteria, and current step.
16	ManageInnovationExperiment	Venture/pilot workflow for hypothesis, assumptions, test method, success thresholds, milestones, observations, learning, and go/pivot/stop recommendation. Evidence updates confidence but never applies a recommendation automatically.	programs-pilots/experiment-studio, preserving pilot, experiment draft, evidence, and recommendation state.
17	ExploreAgentCapabilities	Business-language explorer with search, categories, realistic prompts, and safe previews of all 16 operational tools. Preview cannot confirm actions.	Isolated education/capabilities gallery.

Distinct ownership

• Submission owns creation; business case owns financial shaping after intake.
• Challenge Studio owns strategic demand before ideas exist; idea submission owns one proposed response.
• Personal innovation answers individual status; portfolio exploration answers organization allocation.
• The queue owns workload triage; gate review owns stage evidence; funding review owns capital allocation.
• Budget tracking answers finance stewardship; funding review answers one consequential allocation decision.
• Experiment Studio owns learning after funding; value tracking owns benefits after implementation or scale.
• Growth answers whether the innovation system is gaining momentum; geography answers where participation and conversion are uneven; leadership health reconciles the executive portfolio position.
• Value tracking owns accountability; recognition owns the human celebration resulting from proven impact.
• The brief turns governed data into communication; it does not duplicate analytical exploration.

Role coverage

Role	Primary inline tools	Value shown
Innovator / collaborator	SubmitInnovationIdea, GetMyInnovation, BuildIdeaBusinessCase, CelebrateInnovationImpact	Create, strengthen, follow, and celebrate ideas.
Innovation program manager / challenge sponsor	LaunchInnovationChallenge, GetInnovationGrowth, ExploreGlobalInnovation	Direct participation toward strategic problems and improve program reach and throughput.
Reviewer / gate committee	GetInnovationReviewQueue, ReviewIdeaGate	Prioritize review work and make evidence-based gate decisions.
Finance / investment committee	ReviewInnovationFunding, TrackInnovationBudget	Test a funding decision and steward the full innovation budget.
Venture / pilot lead	ManageInnovationExperiment, TrackInnovationValue	Test assumptions, recommend go/pivot/stop, and remain accountable for outcomes.
Executive leadership / strategy	ExploreInnovationPortfolio, GetInnovationPortfolioHealth, TrackInnovationValue, GenerateInnovationBrief	Balance the portfolio, understand health, verify outcomes, and communicate decisions.

Dynamic inline UX contract

Every inline component is a compact business application, not an Adaptive Card imitation.

• One shared header shows Zava Innovation Hub, the literal action title, and one top-right View in full screen action when the host supports it.
• Prompt properties initialize selected records, filters, comparisons, or editable drafts silently.
• Information views expose meaningful controls and selected detail. Submit and review views expose complete guarded workflows directly inline.
• Composition may change by bounded state: submission moves from canvas to score review; portfolio can switch analytical forms; funding moves from evidence to scenario to confirmation.
• Every visible control changes data, geometry, calculation, selection, draft, validation, or state.
• Charts pair with exact values and a View as table alternative. Status never relies on color alone.
• Inline remains useful at approximately 340 px and polished at approximately 760 px.

Full-screen model

Full screen is one shared five-lens application shell, plus an isolated capability gallery. These are durable work areas in one Copilot application, not five separate applications, and Expand never opens a generic home page merely because a route exists.

• Desktop/keynote uses a restrained vertical rail for My Innovation, Programs & Pilots, Reviews & Gates, Investment, and Enterprise Insights, prioritizing analytical width.
• Narrow widths use an accessible selector or drawer while preserving lens identity, order, state, and focus behavior.
• The product bar contains identity, scope, search, notifications, session settings, persona, and Back to conversation.
• Each lens has a useful default dashboard. A deep-linked inline component becomes the focused module within it and receives keyboard focus.

Profile-aware shell contract

The host-resolved user initializes one demo profile and its owning default lens. The profile changes default scope, priority metrics, queue/action visibility, accountable-person context, and suggested operations; it does not merely change the avatar or greeting. All profiles may navigate to other lenses for the keynote, but role-sensitive mock actions become read-only or require an explicit persona switch. This is demo behavior, not a claim of production authorization; live role and permission enforcement is a deferred integration requirement.

Demo profile	Default lens	First useful full-screen state	Primary handoff
Megan Bowen, innovator	My Innovation	Her ideas, challenge invitations, feedback, next gate, business-case readiness, and recognition.	Submitted idea enters Diego's review queue.
Johanna Lorenz, program/pilot lead	Programs & Pilots	Active challenges, participation reach, funnel growth, experiments needing evidence, and learning decisions.	Challenge creates Megan's opportunity; funded idea becomes an experiment.
Diego Siciliani, reviewer	Reviews & Gates	Pending gate queue, due pressure, evidence gaps, conflicts, and current comparison cohort.	Advanced idea enters Miriam's funding queue.
Miriam Graham, finance partner	Investment	Available/committed/spent/forecast budget, horizon balance, requests awaiting committee action, and decision exposure.	Confirmed funding creates Johanna's pilot envelope and updates Joni's portfolio.
Joni Sherman, executive sponsor	Enterprise Insights	Leadership health, living funnel, strategic balance, growth, geography, value realization, and top exceptions.	Portfolio evidence drives a new challenge, investment action, brief, or recognition.

Every profile has a compact inline demo and a full-screen continuation. Session receipts synchronize the handoffs so switching profiles reveals the prior person's confirmed action without persisting to a tenant system.

My Innovation

The first viewport shows Megan, three living-funnel idea journeys, and the most important next action. Supporting regions coordinate feedback/supporters, upcoming gates, business-case readiness, and recognition. Operations include submit, continue a business case, respond to changes, collaborate, and celebrate impact. Recognition uses illustrated badges and an evidence-grounded share composition rather than a generic points leaderboard.

Challenge invitations and relevant prompts may appear as contextual opportunities, but program-wide challenge design belongs to Programs & Pilots rather than the personal dashboard.

Programs & Pilots

Johanna's dashboard connects strategic demand to validated learning. Its default view combines active challenges, participation reach, submission/conversion growth, regional inclusion, and experiments that need evidence or a recommendation. Challenge Studio focuses outcome, audience, criteria, timeline, and launch review. Experiment Studio focuses hypotheses, thresholds, observations, confidence, and a guarded go/pivot/stop recommendation. Growth and geography expand into coordinated analytical routes rather than competing with the executive command center.

Reviews & Gates

The default dashboard coordinates incoming submissions through Screening, Business case, and Pilot phase buckets with live action/approved/declined/sent-back counts. Review opens one submission at a time with evidence completeness, strategic fit, expected value, strengths, gaps, and accountable owner. Approve advances it, Send back requests named evidence before resubmission, and Decline closes it with a recorded rationale. Every path uses consequence review, explicit confirmation, a semantic session receipt, and an updated queue. Wide layouts show queue plus phase guidance; mobile becomes sequential.

Investment

The Finance Partner dashboard reconciles available budget, committed funding, approved but unspent capital, actual spend, forecast, and horizon allocation. Its main bridge and forecast answer where the budget moved and what remains. Funding Committee is the immersive decision route with business case, cash flow, before/after horizon balance, amount scenario, committee progress, notes, and decision. Expanding budget tracking preserves period, horizon, scenario, and selected variance; expanding a funding review preserves the current request and proposed amount.

Confirmed funding exposes the pilot envelope and expected outcome, then hands execution to Johanna's Programs & Pilots workspace. Investment does not become a generic project-management workspace.

Enterprise Insights

The keynote centerpiece is the proportional living stage-gate funnel. Coordinated regions include impact-vs-effort bubbles, growth trends, a world participation map, horizon balance, theme treemap, leadership health, value realization, and top outcomes. One selected idea, cohort, or location coordinates marks, legend, exact-value rail, and detail. Period, horizon, theme, region, and stage rebuild compatible models. The default leadership dashboard stays concise; focused growth and geography routes gain the space required for cohort comparison and map detail in Programs & Pilots. Enterprise Insights retains read-only leadership rollups and deep-links to those operational routes with period, region, cohort, and selected evidence preserved.

The dashboard closes with Top performing ideas by value realized: a ranked, people-centered view of idea, contributor, verified value, ROI, current phase, and earned impact badge. Leaders can acknowledge an outcome directly; the React experience creates a session receipt and prepares the recognition story. This is deliberately not a popularity leaderboard. Rank requires measured value, evidence confidence, and named contribution; raw submission volume, reactions, and votes do not create status.

Joni can initiate a new challenge or inspect an experiment from portfolio evidence, but the typed destination moves to Johanna's Programs & Pilots lens with the selected objective, region, idea, or exception preserved. Enterprise Insights remains leadership-oriented rather than becoming campaign or pilot administration.

Profile demo paths

Demo	Inline proof	Expand destination and added value
Innovator	Megan submits or checks an idea; fields and scores react to intent.	My Innovation preserves draft/idea and adds feedback, milestones, case readiness, and recognition.
Program lead	Johanna frames a challenge, inspects global reach, or records experiment evidence.	Programs & Pilots preserves challenge/region/experiment and adds program context, cohorts, and learning decisions.
Reviewer	Diego opens his queue or reviews one gate.	Reviews & Gates preserves filters/cohort/draft and adds side-by-side evidence and reviewer calibration.
Finance	Miriam inspects budget or models one funding request.	Investment preserves period/scenario/request and adds reconciled budget, horizon impact, notes, and committee decision.
Executive	Joni asks for health, landscape, value, or a brief.	Enterprise Insights preserves metric/filter/selection and adds coordinated funnel, growth, geography, balance, value, and exceptions.

Capability gallery

The isolated gallery provides category navigation, search, audience/operation filters, realistic prompts, Previous/Next, and a featured tour. Read-only previews use the shared experience router but stop before consequential confirmation.

Visual direction

The PNGs in assets are composition references, not implementation screenshots. The result should feel like a modern venture studio and investment committee, not a suggestion box or generic dashboard template.

• Canvas: neutral Fluent surfaces and generous unframed analytical areas. Cards are reserved for ideas, people, decisions, modals, and bounded tools; never nest cards.
• Color: deep grape is the governance anchor and marigold is creative energy. A restrained bridge tone may support ordered analytical ramps; Fluent semantic colors retain status meaning.
• Typography: Segoe UI Variable, compact operational headings, and tabular numerals. Funding, value, and scores become display moments only where hierarchy warrants it.
• Imagery: the 11 bundled reference portraits for named people, plus original illustrations for themes, badges, empty states, and the agent mark. No runtime media fetch, robot chrome, or generic AI sparkles. Portrait redistribution remains gated on recorded source rights.
• Motion: 200-320 ms gate-opening, bounded chart entrance, cross-highlighting, and settled funding updates. Reduced motion reveals final state immediately; no perpetual animation or fake waiting.
• Signature visuals: living funnel at three scales, funding consequence, impact constellation, share-ready recognition artifact, and the company impact-recognition board grounded in verified people, phases, ROI, and realized outcomes.

Visualization decisions

Model first, renderer second. Each visual consumes a pure typed model with stable IDs, labels, exact and formatted values, relationships, semantic colors, geometry inputs, selection, legend, and fallback rows.

Question	Visual	Direction
Where is an idea or cohort?	Living funnel / journey	React SVG plus Fluent DOM evidence.
How strong is the idea?	Five-axis radar and criterion bars	React SVG with exact labels.
Which ideas combine impact and feasibility?	Bubble matrix	React SVG; evaluate headless d3-scale only if it materially helps.
Is investment aligned?	Donut/stacked balance with target	React SVG plus exact legend.
Where is funding concentrated?	Theme treemap	React SVG; evaluate d3-hierarchy in a measured spike.
How did budget move?	Funding waterfall	React SVG with connectors and totals.
Are pilots realizing value?	Milestone horizon and value curve	Fluent DOM timeline plus React SVG line.
Is the innovation program growing sustainably?	Layered line/area chart with cohort bands and conversion overlay	React SVG using headless d3-array, d3-scale, and d3-shape when the measured spike passes.
Where is participation and conversion uneven?	World choropleth/proportional-symbol map plus ranked regional bars	React SVG using local boundaries with d3-geo and topojson-client; never fetch map tiles at runtime.
What is the current and forecast budget position?	Funding bridge plus burn/forecast line and horizon allocation	React SVG using focused D3 array/scale/shape math where it replaces hand-rolled stacking and paths.
Is the portfolio healthy enough for leadership action?	Health ribbon, funnel velocity, balance, concentration, and exception trend	Fluent DOM plus coordinated React SVG; selection opens exact evidence rather than another generic dashboard.

Babylon is not in the default plan. A measured spike may propose one dimensional portfolio scene only if depth materially improves a decision and complete 2D, accessibility, performance, fallback, and bundle evidence remain available. Do not use 3D merely as spectacle.

Coherent mock data

The deterministic offline graph includes strategic objectives, eight active/completed challenges, at least 120 ideas across every stage; 100 employees across AMER, EMEA, APAC, and LATAM with country and business-unit attribution; six uneven strategic themes; at least 24 months of submission/contributor/gate history; budget baselines, allocations, commitments, actuals, forecasts, and variance drivers; business cases and cash flows; gate reviews; experiment hypotheses, thresholds, observations, evidence, learning, and recommendations; full/partial/deferred/rejected funding; 15 pilots; six quarters of history; value measurements; feedback; votes; recognition evidence; and mock share receipts. Smart Onboarding Journey stays coherent across all of these records.

Dates derive from relative offsets and one invocation clock. Confirmed demo actions append to a guarded session-only overlay; Reset restores immutable seeds. Currency uses Intl.NumberFormat, defaults to USD, and remains locale-ready. Production boundaries are typed service interfaces, pure mappers/calculators, workflow transitions, and the session action store. Live adapters remain deferred.

Architecture and routing

• One typed intent catalog owns names, GUIDs, operations, schemas, positive/negative routing boundaries, previews, education metadata, and full-screen destinations. One ordered starter configuration owns visible prompts and their expected inline-component targets.
• Each final component receives an immutable Yeoman-generated scaffold when its phase starts. The existing Innovation placeholder is retired, not renamed or repurposed.
• One shared React 17 + Fluent UI v9 + Griffel host renders into the component ownerDocument.
• Components share one measured SPFx bundle to avoid repeating React, Fluent, media, and domain data.
• Deterministic signatures separate fresh invocation from passive rerender. Transient selection, filters, drafts, and review state survive Expand; confirmed actions alone enter the session overlay.
• Full-screen destinations are catalog metadata, never component-name conditionals.
• The showcase makes no runtime network calls.

Conversation starters

Use exactly six starters. Each targets a distinct inline tool across personal submission, portfolio insight, idea approval, and capability discovery. Starter 6 always opens the capability explorer.

#	Title	Starter	Expected inline component
1	Submit an idea	Submit an idea to reduce new-hire onboarding time by half.	SubmitInnovationIdea
2	Global innovation	Compare innovation participation and conversion across regions.	ExploreGlobalInnovation
3	Portfolio funnel	Show stage conversion in our innovation portfolio funnel.	ExploreInnovationPortfolio
4	Portfolio health	Show which innovation portfolio health measures are outside target.	GetInnovationPortfolioHealth
5	Approve an idea	Review Smart Onboarding Journey for gate approval.	ReviewIdeaGate
6	Explore capabilities	Explore what this agent can do.	ExploreAgentCapabilities

Every request selects one primary tool. Related evidence appears inside that component rather than invoking score, sponsor, ROI, and balance tools in parallel. Local collision tests validate authored boundaries; authenticated Workbench rehearsal proves actual model routing.

Keynote flow

1. Turn the strategic goal to halve ramp time into a measurable Future of Onboarding challenge; preview audience, criteria, expected reach, and portfolio fit before mock launch.
2. Submit Megan's Smart Onboarding Journey response; edit the outcome and watch similarity, score, and readiness react.
3. Review it against its cohort and advance it through a guarded gate decision.
4. Adjust full to partial funding and show portfolio balance, value, and milestone scope respond.
5. Continue into the pilot experiment: one observation misses the adoption threshold, changing the recommendation from Go to Pivot without automatically making the decision.
6. Expand with exact context into Enterprise Insights; show funnel, growth, and leadership health update.
7. Close with measured onboarding value and an evidence-grounded praise card for Megan and collaborators.

The longer business demo adds business-case sensitivity, reviewer calibration, value tracking, executive brief generation, and capability discovery. The technical walkthrough shows catalog, mock graph, chart models, workflows, host state, exact continuation, accessibility, evidence, and package audits.

Designer and demo review

• Designer review guide
• 4-minute keynote
• 10-minute business demo
• 5-minute technical demo
• Prompt and routing matrix
• Machine-readable visual evidence
• Publication gallery evidence
• Authenticated Workbench evidence
• Production package evidence

Build the local review harness with npm run build:visual, then serve temp/visual-harness over a local HTTP server. Query parameters select intent, mode=fullscreen, and theme=dark. The design guide lists the primary screenshots, interaction paths, accessibility states, and feedback format.

Prerequisites and setup

• Node.js 22.14 through 22.x and npm 10.x.
• A SharePoint tenant with Copilot Component Workbench access for authenticated host validation.
• The trusted SPFx development certificate for localhost Workbench testing.

npm ci
npx playwright install chromium
npm run validate
npm run build

npm run build validates the 17 manifests and six starters, agent icons, media provenance, byte-exact routing documentation, current-source gallery screenshots, production tests, generated API plugin, and final package contents. It emits the deployable package at sharepoint/solution/zava-innovation-portfolio.sppkg and release hashes in assets/release-evidence.json.

For deterministic UX review, run npm run capture:visual. For authenticated Workbench development, set SPFX_SERVE_TENANT_DOMAIN, run npm start -- --nobrowser, and load the emitted debugManifestsFile query string in the tenant Copilot Workbench. Direct component selection proves host rendering and interaction; natural-language model routing must be rehearsed separately.

Validation and limitations

• All 17 inline components render in authenticated Workbench with owner-document Fluent styling, light/dark theme updates, zero horizontal overflow, and representative workflow interaction.
• The local publication matrix contains 11 source-hash-backed captures with zero runtime, image, or overflow failures. Broader deterministic evidence covers all inline defaults and five workspaces.
• The production dependency graph has zero known vulnerabilities. Nine moderate findings remain in the pinned SPFx/Heft development toolchain; npm offers only a breaking downgrade that does not support this Copilot Component preview.
• Live Dataverse, SharePoint, Graph, Fabric, finance, durable workflow, and notification adapters are intentionally deferred. Mock actions remain session-local and do not write tenant data.
• Natural-language starter routing, tenant screen-reader coverage, and portrait redistribution approval remain external release prerequisites and are not claimed by the local build.

Scope and definition of done

The current showcase includes 16 operational inline components, one education component, five shared full-screen lenses, one isolated gallery, deterministic mock workflows, full responsive/themed UX, and catalog/routing/asset/visual/plugin/package automation. Live Dataverse, SharePoint, Graph, Teams praise, Fabric, finance, durable workflow, authorization, audit, provisioning, and notifications are deferred.

The release candidate has 17 immutable identities, distinct inline defaults, guarded mock actions, responsive workspaces, deterministic visuals, a safe capability explorer, authenticated direct-selection Workbench evidence, and an audited .sppkg with no stale bundles or duplicate media. Final publication still requires the external prerequisites listed above.

---

This implementation supersedes the earlier 30-inline-component, interim 10-tool, and 14-tool proposals. The approved 16 operational tools plus one capability explorer are the current release catalog.