Your Workspace

Explore with The Librarian

Your workspace's AI research assistant — ask it anything about your users, your insights, or your product, and it cites the evidence.

The Librarian is Seena's in-workspace research assistant. You ask it questions in plain language, it answers with evidence from your actual user data.

Ask it "Why are users dropping off at checkout?" and it will answer with insights, quotes, and specific session events that support the answer. Every claim is cited.

What the Librarian draws from

The Librarian sits on top of everything Seena has collected for your workspace:

  • Briefings — the daily and weekly narratives.
  • Insights — the atomic findings, clustered.
  • Session nodes — individual visitor sessions with their events and interview transcripts.
  • Product Knowledge — as users visit your site, Seena explores the pages they are browsing, learns them, and caches those learnings until you update your product. That way Seena always has an updated knowledge about what your product is.
  • Your preferences — the more you chat with Seena the more it learns about you, tailors, and improves its output to match serve you best.

When you ask a question, the Librarian retrieves across all tiers and synthesizes an answer. For broad strategic questions, it leans on briefings and insight clusters. For specific questions about a user or a session, it drills into the raw session data. And when it learns something new about you, it adds it to the key facts sections that is easily accessible to you on the right hand side to view, edit, or delete.

Shared vs. personal facts

The Librarian remembers two kinds of facts:

  • Shared facts — about your product, your users, your company, or decisions the team has made. These are visible to everyone in the workspace. Example: "Pricing changed from $49 to $59 on March 1."
  • Personal facts — about you specifically. Your preferences, your current focus, how you like the Librarian to answer. These are private to you; nobody else sees them.

The Librarian decides which category a new fact belongs in automatically based on what you tell it. You can review and correct these in Key Facts panel.

Scope

The Librarian is workspace-wide. Even if you've assigned team members to specific products or sites in your workspace, the Librarian can see everything. This is deliberate — cross-product patterns are often where the most valuable insights live.

What it's good at

  • Synthesis across many sessions. "What are the top three frustrations on our onboarding this week?"
  • Specific user deep-dives. "Tell me everything about user id 1234's last five sessions."
  • Trend questions. "Has sentiment on the checkout flow improved since we redesigned it?"
  • Comparison questions. "What do Pro users complain about that Max users don't?"
  • Quote-finding. "Find me a clean quote about pricing confusion I can share with the team."

What it's less good at

  • Questions whose answer isn't in your data. The Librarian won't speculate. If no sessions touched your pricing page, it won't pretend to have opinions about your pricing.
  • Highly specific factual recall from a single interview. It'll point you at the interview, but if you need the exact 7-second clip, you'll need to open the interview itself.

Cost

Librarian conversations are billed to whoever is chatting, from their own usage balance. Simple questions are a tiny line item; deep multi-tier retrievals across briefings, clusters, and session data cost a bit more.

What to read next