# Quickstart

# Quickstart

This quickstart takes you through the smallest “end-to-end loop” that demonstrates the Curiosity Workspace value chain:

1. Define a graph schema
2. Ingest data (via connector or integration)
3. Query the graph
4. Make the data searchable
5. Add AI (embeddings / semantic search) where it helps

# Step 0: Create (or choose) a workspace

• Create a new workspace environment (or reuse an existing dev workspace).
• Confirm you can open the Management/Admin UI.

# Step 1: Model your data as a graph

Start with:

• Node types: what are the “things” in your domain? (e.g., Customer, Ticket, Product)
• Edge types: how do those things relate? (e.g., Opened, Uses, RelatedTo)
• Keys: which fields uniquely identify a node (stable IDs)?

Good schemas optimize for:

• Navigation: “show me related things”
• Filtering: “limit results to this entity/time/status”
• Searchability: “index the fields users search”

For guidance, see Core Concepts → Graph Model and Data Integration → Schema Design.

# Step 2: Ingest data

You have two common ingestion patterns:

• Connectors: programmatic ingestion that maps source records into nodes and edges.
• Pipelines / integrations: configured ingestion flows that pull from external systems.

Start with Data Integration → Connectors.

# Step 3: Validate in the graph (query + explore)

After ingestion, validate:

• counts per node type
• expected edges exist
• keys are unique (no duplication)

Use the query interfaces described in Reference → Graph Query Language.

# Step 4: Enable search

Pick which node types and fields are searchable:

• Text search fields: titles, summaries, identifiers, descriptions
• Filter facets: properties (status, type) and related-entity facets (manufacturer, customer)

See Search → Text Search.

# Step 5: Add AI search (embeddings) where it improves recall

Add embeddings to fields where keyword matching is insufficient (long descriptions, conversations, notes).

See:

• Search → Vector Search
• NLP → Embeddings

# Step 6 (optional): Extend with endpoints and custom UI

Once the basics work, you can:

• add custom endpoints for business logic (aggregation, similarity, orchestration)
• build a custom front-end tailored to your users (workflows, dashboards)

See APIs & Extensibility.

# Next steps

• Understand how the platform fits together: Core Concepts → Architecture
• Design production-grade schemas: Best Practices → Graph Design Patterns