Intelligence
LealUp's AI features, what each one does, when to use it, and which to pick for the job.
Intelligence groups LealUp's AI capabilities. Each one is built for a different moment in a CSM's day: researching before a meeting, drafting a reply, spotting risk, synthesizing a conversation, or asking an open question against real context.
It is not one magic assistant. They are specific agents with a declared scope, all fed with your workspace's data (not other customers' data, and not invented public data).
Which to use and when
Analytics
Aggregate dashboards and metrics across the book. For leadership and CS Ops.
RAG Search
Search a customer's whole conversation history and get answers with citations.
Briefings
Automatic pre-meeting preparation, with customer context and open topics.
Sentiment
Sentiment detection over emails and messages, to see problems coming.
Ask this customer
Open question with a specific customer's context (or a specific contact's).
Luminance Chat
Cross-portfolio chat for executive questions about the whole book.
Quick decision matrix
| I want to… | Use |
|---|---|
| See an aggregate metric (average health, this month's renewals) | Analytics |
| Know what the customer said about a specific topic | RAG Search |
| Get ready for a meeting in 30 seconds | Briefings |
| Detect whether a customer is unhappy | Sentiment |
| Answer a question about one customer's specific context | Ask this customer |
| Explore across the book: "which customers are asking for X" | Luminance Chat |
Design principles
- Every agent has a declared scope. It will not answer outside its context. A briefing does not answer open questions; RAG search does not draft replies.
- Answers cite their source. Whenever an agent uses specific content (an email, a WhatsApp message, a note), you see where it came from and can open it.
- Your data stays in your workspace. Embeddings are stored tenant-scoped; LLM queries are sent without identifiers linking to other accounts.
- Alfred is not part of this. Alfred is the product guide (it teaches you to use LealUp). The features in this section work with your customer data. Do not confuse them.
General limitations
- The answer's language follows the source data's language; if your conversations are in Spanish, the answers will be in Spanish.
- Recent data takes around 5 minutes to be indexed (RAG).
Last updated: 2026-04-21.