Llumen platform one-pager — partner documentation for 7 core features with design system styling

The last mile of
intelligence

Llumen is a governance-first ecosystem that consolidates an organization's data into a single platform — enabling teams to build, compose, and share decision-ready intelligence with leadership.

Platform overview
Everything connects through
the governance layer

Llumen unifies ingestion, insight design, AI, security, and operations in one governed stack — so teams ship charts, maps, and narratives from the same trusted data without re-wiring pipelines for every new question.

01 — Ingestion Data Sources Integration (API, ETL, Upload Transformation, Generative Models)

The intake layer where APIs, scheduled ETL, governed file uploads, and generative enrichment meet the same validation rules — normalizing schemas, lineage, and refresh cadence before anything reaches an insight. Connect once, reuse everywhere, and keep every transformation auditable for compliance and downstream trust.

Product UI — product walkthrough or architecture diagram

API & streaming connectors
First-class connectors with rate-aware retries, schema drift detection, and health signals in one operations view
ETL with lineage
Versioned jobs, dependency graphs, and replayable windows so analysts know exactly which build produced each metric
Upload transformation
Ingest spreadsheets and documents through validation templates — reject bad rows early instead of debugging dashboards later
Generative enrichment
Optional model-assisted tagging and summarization that stays inside RBAC, with human approval gates before writes land in core tables
02 — Insights Insight Builder for Charts

A purpose-built surface for KPIs, time series, and comparisons — where every series binds to governed measures, color semantics stay on-brand, and drill paths respect row-level security by default. Design the story once, publish to dashboards, briefings, and narratives without rebuilding the logic in three different tools.

Product UI — screen recording of chart builder

Measure library
Central definitions for metrics, filters, and grain — change a formula once and every dependent chart inherits the update
Interaction presets
Cross-filtering, brush-linked views, and guided drill templates that stay consistent across workspaces
Export-ready layouts
Pixel-aware spacing and safe zones so leadership-ready exports match what you see in the builder
Performance guardrails
Query budgets, sampling strategies, and cache hints surfaced before publish so heavy charts never surprise production
03 — Spatial Insight Builder for Maps

Spatial insight design with the same governance backbone as charts — layers, legends, and camera paths stay data-bound, not hand-drawn. Build once for operations rooms and executive walkthroughs alike, with transitions that read as narrative beats instead of disconnected screenshots.

Product UI — map interaction capture or static hero

Layer stacks
Vector, raster, and tabular overlays with shared filters and synchronized time sliders across the stack
Geospatial RBAC
Region-aware entitlements so sensitive assets never render outside approved geographies or roles
Annotation & callouts
Persistent markers tied to incidents or KPI thresholds — always linked back to the originating dataset
Export to narratives
Push the same map state into slide-ready frames without losing interactivity metadata for the briefing room
04 — Personalization Personalized leadership landing page

Every leader gets their own home screen, curated around their role and priorities — briefings, followed topics, suggested narratives, and highlights assembled into a single view that makes the most important information immediately visible.

Product UI — role-based home and suggested narratives

Priority briefings
Urgent push-delivered reports flagged directly in the inbox
Tracked KPIs
Personally followed topics and indicators visible at a glance
Role-curated view
Highlights and narratives surfaced by the user's team and role
AI-suggested narratives
Intelligent recommendations based on focus areas and recent activity
05 — Communication Narratives

Multi-slide presentations built from governed components — replacing static slide decks with a format that stays connected to live data. Maps inside narratives are fully interactive, with cinematic camera transitions guiding leadership through a spatial story.

Product UI — multi-slide story with interactive maps

Governed components
Compose slides from any combination of charts, KPIs, and maps
Cinematic transitions
Camera animations between slides for immersive spatial storytelling
Interactive spatial layers
Tooltips and data drill-downs directly on map components
Always live
Connected to real-time data — no manual refresh or export needed
06 — Urgency Briefings

Condensed, mobile-first narratives built for high-urgency moments. Pushed directly to leadership and designed to be consumed immediately on a phone — in a story format that requires no context-switching or setup.

Product UI — mobile-first urgent narrative

Story-style format
Swipe-through slides optimized for phone screens
Push-delivered
Flagged high-priority in leadership's inbox on arrival
Condensed narratives
Distilled from full narratives while staying connected to live data
60-second reads
Structured for immediate comprehension under time pressure
07 — Monitoring Dashboards

Components from the governance layer arranged in a modular grid — delivering a command-and-control experience with Llumen's data integrity underneath every tile. Charts, KPIs, and maps coexist in any configuration.

Product UI — modular grid of KPIs, charts, and maps

Modular grid layout
Drag-and-drop arrangement and free resizing of tiles
Real-time refresh
All tiles update simultaneously from live data streams
Mixed component types
Charts, KPIs, and maps in a single unified view
RBAC access control
Shareable with the same permission framework as all Llumen surfaces
08 — Automation Workflows

A visual node-based editor for building automation pipelines without code. Connect data triggers, conditions, simulations, and delivery actions — and when conditions fire, Llumen automatically generates and dispatches a narrative to the right people.

Product UI — visual automation and dispatch

Visual node editor
Build multi-step automation logic without writing a line of code
Threshold & schedule triggers
Fire on data conditions or recurring time-based schedules
Simulation pipelines
Automated scenario testing that converts to narratives on completion
Auto-dispatch narratives
Generates and routes briefings to leadership the moment a threshold is met
09 — Intelligence Llumen AI

The in-platform agent that reasons over governed components instead of scraping whatever a browser session can see — answering in natural language, citing the charts and maps it used, and refusing requests that violate policy. It is the fastest path from question to evidence without opening a separate “AI tool” that bypasses your controls.

Product UI — conversational UI or agent flow

Component-grounded answers
Responses assemble from approved measures, narratives, and map states — not from opaque web fetches
Task playbooks
Saved flows for recurring investigations so teams repeat the same quality bar every Monday morning
Guardrailed actions
Writes, shares, and publishes require explicit scopes — the agent proposes, humans or policy approve
Audit trail
Prompt, retrieval set, and output hashes retained for security reviews without slowing day-to-day use
10 — Models LLM Agnostic Integration (choose your model or bring your own)

Model choice stays with you — swap vendors, regions, or privately hosted weights without rewriting every workflow. Route traffic by sensitivity tier, pin cost ceilings per workspace, and keep key material in your vault while Llumen orchestrates prompts, retrieval, and evaluation harnesses in one place.

Product UI — provider switcher or BYOK diagram

Multi-provider routing
Policy-driven selection between hosted, partner, and on-prem endpoints with failover that preserves queue integrity
Bring-your-own keys
Tenant-scoped secrets, rotation hooks, and spend telemetry surfaced next to every model family you enable
Evaluation suites
Regression prompts, golden answer sets, and latency SLOs tracked per release so upgrades are deliberate
Prompt & tool registry
Versioned templates and allowed tool lists — no shadow prompts running outside the catalog you approved
11 — Security Security Access - Bring your own keys, single sign on (Azure, google, any other key)

Identity and secrets land where your security team already operates — SAML/OIDC with Azure AD, Google Workspace, or any standards-compliant IdP, plus customer-managed keys for the data plane. Fine-grained workspace roles inherit from your directory groups, and every API call carries the same proof-of-membership your auditors expect.

Product UI — SSO flow or key vault integration

Enterprise SSO
JIT provisioning, group mapping, and session policies aligned to your IdP — no duplicate user stores to reconcile
Customer-managed keys
Envelope encryption with your KMS so revocation is instantaneous without waiting on vendor queues
Service accounts & scopes
Machine identities with least-privilege bundles and expiring tokens for automation that cannot sprawl silently
Continuous posture signals
Exportable access reviews, anomaly hints, and integration hooks for SIEM without duplicating raw telemetry off-platform
12 — Operations Reporting and Monitoring Access reports

Operational clarity without another silo — scheduled PDFs, CSV extracts, and live monitoring tiles draw from the same governed definitions your builders use, so “the report number” and “the dashboard number” cannot diverge. Watch delivery health, subscriber acknowledgements, and data freshness in one surface tuned for platform owners.

Product UI — report catalog or schedule UI

Report catalog
Searchable library with ownership tags, deprecation notices, and lineage back to the components each report consumes
Schedules & handoff
Timezone-aware distribution lists, encrypted attachments, and webhook fan-out for systems that need the signal
SLA monitors
Latency, failure rate, and backlog depth per pipeline with paging rules tied to your on-call rotation
Access attestations
Periodic recertification queues so managers confirm who still needs sensitive operational packs
13 — Resilience Backup and Restoring

Predictable recovery when mistakes or incidents happen — object-level backups, workspace-scoped restores, and integrity checks that run before a dataset is marked live again. Retention windows follow your compliance tier, with cold storage routing and legal hold flags that do not break the nightly cadence for everyone else.

Product UI — restore walkthrough or retention policy UI

Point-in-time snapshots
Coordinated captures across metadata and blobs so restores never leave dangling foreign keys
Workspace rollback
Undo a bad publish or migration for a single team without rewinding the entire tenant
Integrity verification
Checksums and row-count reconciliations after every restore — failures bubble to operators with suggested fixes
Geo-redundant copies
Region policies you choose, with bandwidth shaping so large artifacts do not starve interactive traffic
14 — Adoption Onboarding Wizard

A guided first-run that connects identity, data sources, sample insights, and notification defaults in one sitting — so new regions or business units reach “first governed dashboard” in hours, not weeks. Each step explains the downstream effect, with skip-safe defaults that still keep you inside policy when teams are in a hurry.

Product UI — wizard steps screen capture

Identity & workspace setup
IdP mapping, role templates, and starter sandboxes that mirror production guardrails without copying sensitive data
Data source checklist
Credential tests, schema previews, and recommended refresh windows before anything hits a builder canvas
Sample narratives
Pre-built stories that demonstrate cross-component behavior — swap in your measures when ready
Success milestones
Track completion per stakeholder with nudges only when someone is blocked, not on a noisy daily blast
15 — Brand Themes

Brand and accessibility without forked code — semantic tokens for surfaces, typography, and chart palettes propagate from a single theme pack so every new component inherits contrast-safe defaults. Preview changes per workspace, schedule rollouts, and roll back instantly if a color regression slips past design QA.

Create Theme — palette colors, background, and logo options

Product UI — light/dark theme comparison

Token-driven styling
Radii, spacing, and elevation scales shared across builders, dashboards, and exported decks
Light / dark parity
Paired palettes with tested contrast pairs — no one-off hex edits that break when you add a new component type
Workspace overrides
Sub-brand or regional skins that still inherit global safety rails for data visualization
Import from design tools
Bring Figma variables or CSS packages, map them once, and keep drift detection on every deploy
16 — Administration Manage Users and Workspaces

Administration that scales with the org — invite flows, delegated ownership, and workspace boundaries that mirror how you segment data today, not how a generic SaaS assumed you would. Bulk operations carry dry-run previews, and destructive actions require stepping stones your security team configures centrally.

User management — users list with roles, status, and last active

Product UI — admin console or role matrix

Role templates
Curated bundles for analyst, curator, and executive personas — extend without inventing net-new permission math each time
Workspace lifecycle
Archive, merge, or export a workspace with dependency checks so shared components are never orphaned silently
Audit-friendly invites
Time-bound links, mandatory SSO domains, and automatic deprovisioning when HR signals leave
Usage visibility
Seat consumption, builder hours, and storage rollups per workspace for finance and platform owners
17 — Context Global Filters

A shared context bar that travels with the user — time range, geography, business unit, and scenario selections propagate to every governed surface that opts in, without fragile URL hacks or duplicated filter SQL. Saved contexts become shareable links that hydrate charts, maps, and narratives identically for everyone in the review.

Add component — filters selection on map

Product UI — filter bar across two surfaces

Cross-surface sync
Opt-in components subscribe to the same context object — change once, watch dashboards and briefings align
Saved contexts
Named slices for "Q3 LATAM" or "incident 4721" with permissions on who may publish or fork them
URL & API parity
Deep links and headless consumers read the same filter envelope — no divergent parameter vocabularies
Guarded combinations
Block impossible filter tuples before query time with plain-language explanations for the analyst