Start Here

Research Librarian: Getting Started

Start using Research Librarian with a controlled example.

ProductResearch LibrarianVerified version7.0.0ComponentCitation Handoffs, Endpoint Diagnostics, Evidence Cards, Feedback and Ratings, Follow-up Questions, Human Review Boundary, Index Health, Intent Classification, Knowledge Library Retrieval, Product Routing, Question Intake, Research Path Assembly, Saved Journeys, Scope Gate, Source DiscoveryUpdatedJuly 17, 2026Reading time18 minutes
← Back to product guides

Research Librarian · Getting Started

Site-scoped research guidance that routes questions across knowledge, products, sources, and saved journeys.

Audience: product users, administrators, reviewers, and integrators.

Last verified version: 7.0.0

Reading time: 18–30 minutes

Status: Published support documentation.

Purpose

This guide provides a safe first run of Research Librarian. It explains the product role, the main feature surface, the synthetic demonstration dataset, how to recognize a valid result, and when human or expert review is required.

Capability overview

  • Question Intake
  • Scope Gate
  • Intent Classification
  • Knowledge Library Retrieval
  • Product Routing
  • Source Discovery
  • Research Path Assembly
  • Follow-up Questions
  • Evidence Cards
  • Citation Handoffs
  • Saved Journeys
  • Feedback and Ratings
  • Index Health
  • Endpoint Diagnostics
  • Human Review Boundary

Prerequisites

  • A modern supported browser or runtime
  • The current installed product and service version
  • A new demonstration workspace
  • No confidential, personal, credential, or production information
  • A named reviewer for consequential output

Sample data

This article uses a fictional Resilient Community Retrofit Program. The files contain one demonstration row for every documented feature, so each capability can be tested without private or production data.

Download CSV feature examples · Download JSON workspace example

Preview

# Feature Sample action Expected result Classification
1 Question Intake ask how a city should compare retrofit options a normalized research request synthetic-demo
2 Scope Gate keep the request within Sustainable Catalyst research and products an in-scope route or clear boundary synthetic-demo
3 Intent Classification classify the request as sourcing, calculation, analysis, or decision support a transparent intent record synthetic-demo
4 Knowledge Library Retrieval retrieve energy-efficiency and decision-method documents ranked library records synthetic-demo
5 Product Routing route calculation to Workbench and synthesis to Decision Studio a product sequence with reasons synthetic-demo

The downloadable files include all 15 feature rows.

Procedure

  1. Open the product from its canonical Sustainable Catalyst page or administrative route.
  2. Confirm the product and backend version before creating data.
  3. Create a synthetic workspace named RESEARCH-LIBRARIAN-RETROFIT-001.
  4. Download the CSV and JSON examples supplied below.
  5. Use the feature matrix to complete the first six feature checkpoints.
  6. Save, close, and reopen the workspace to verify persistence.
  7. Inspect warnings, sources, units, permissions, and lifecycle state.
  8. Export or hand off only after completing the verification checklist.

1. Question Intake

Action: ask how a city should compare retrofit options.

Expected result: a normalized research request.

2. Scope Gate

Action: keep the request within Sustainable Catalyst research and products.

Expected result: an in-scope route or clear boundary.

3. Intent Classification

Action: classify the request as sourcing, calculation, analysis, or decision support.

Expected result: a transparent intent record.

4. Knowledge Library Retrieval

Action: retrieve energy-efficiency and decision-method documents.

Expected result: ranked library records.

5. Product Routing

Action: route calculation to Workbench and synthesis to Decision Studio.

Expected result: a product sequence with reasons.

6. Source Discovery

Action: identify utility, building, climate, and finance source types.

Expected result: a source shortlist with provenance expectations.

Expected output

  • a normalized research request
  • an in-scope route or clear boundary
  • a transparent intent record
  • ranked library records
  • a product sequence with reasons
  • a source shortlist with provenance expectations
  • A saved record that reopens without data loss
  • An export or handoff that identifies product, version, artifact, source, and review state

Verification

  • Every sample value remains attached to its identifier and classification.
  • Warnings remain visible and are not silently treated as approval.
  • The saved workspace reopens in a clean session.
  • The product and version shown in output match the installed release.
  • Another reviewer can understand how the output was produced.

Responsible use

This is a routing and discovery aid, not a general chatbot or final authority; inspect original sources and apply domain judgment.

Scroll to Top