Working with Lumen AI
What Lumen can do, how stage-aware context works, and how to confirm or dismiss AI suggestions.
Lumen is Cursus's AI copilot. It's grounded in your organization's live data — programs, stakeholder groups, readiness scores, network structure — not just general knowledge.
What Lumen can do
Draft communications
Lumen can generate stakeholder communications, change announcements, and leader talking points. Provide context ("draft a message for the Finance group explaining the ERP timeline change") and Lumen produces a draft tailored to that group's current E-C position, the program's change philosophy, and the stage of the program.
Generate change impacts
On any program's Impacts tab, ask Lumen to generate a first draft of change impacts based on the program description. Review and edit before saving. This is faster than starting from a blank form, especially for complex programs with many affected groups.
Coach leaders
In the Leadership Portal, Lumen operates as a coaching assistant. It uses a leader's scoped data — their team's change load, E-C position, and network signals — to provide situational coaching guidance. Prompts are grounded in the organization's configured change philosophy.
Analyze transcripts and notes
Paste meeting notes, town hall transcripts, or feedback summaries into Lumen and ask it to extract themes, flag resistance signals, or summarize sentiment. Results are for practitioner use and are not stored as behavioral data about individuals.
Stage-aware context
Lumen adapts its behavior based on where you are in the application. When you open the Lumen panel on the Impacts page, it knows you're working on impacts. On the E-C Matrix, it defaults to readiness analysis. On an individual program's Settings page, it offers program configuration guidance.
This means you don't need to re-explain context in every message. Lumen uses the current page, the active program, and the data in view as implicit context for every prompt.
How AI suggestions are marked
All Lumen-generated content is displayed with a teal background and left accent border. This visual treatment persists until you confirm the suggestion. Once you confirm, the treatment transitions to indigo — indicating the content is now practitioner-owned, not AI-suggested.
This distinction is intentional. Teal = AI suggested. Indigo = practitioner confirmed.
The confirmation workflow
Lumen never auto-applies suggestions. Every piece of AI-generated content — a communication draft, an impact list, a briefing document — requires an explicit confirmation action before it is saved or acted on.
The typical workflow:
- Ask Lumen to generate or draft something
- Review the teal-highlighted suggestion
- Edit inline if needed
- Click Confirm to accept, or Dismiss to discard
Nothing is sent, saved, or logged until you confirm. This applies to communications, impact records, and any structured output Lumen produces.
Change philosophy awareness
Lumen's language, coaching prompts, and intervention recommendations all adapt to your organization's configured change philosophy. An ADKAR-configured organization will receive responses that reference stages like "Desire" and "Reinforcement." A Bridges-configured organization will hear Lumen speak about the Neutral Zone and transition support.
The philosophy is set by your administrator in Settings → Organization → Change Philosophy.