Ask a question

Type your question in natural language in the input field, then submit. No need to know the query language: the assistant takes care of it.

The examples on this page are based on a fictitious demo environment: brands (Brand) produce models (Model) based on platforms (Platform). Each model contains functional assemblies (Assembly) made of parts (Part), purchased through contracts (Contract) negotiated with suppliers (Supplier).

Example:

Which models are built on the Nova platform?

What the assistant does

The assistant does not translate your question into a single query: it works step by step.

1. It reads the ontology to find the relevant object and relationship types
2. It finds the actual objects matching the names you used
3. It generates a Cypher query using Text2Cypher
4. It queries the graph and presents the result

Progress messages show you the current step.

Result

Depending on your question, the assistant answers with:

• A table — for a list of records
• A chart — when the data lends itself to it
• Text — for a single value, a yes/no or an explanation
• A question — if your request is ambiguous, a form with clickable options is displayed. Pick an option or use Other to type your own

Tables and charts always come with a written explanation.

Result options

• Open graph — opens the result as a graph [TODO: confirm whether it opens in Explore]
• Refresh — reloads the table data
• Explain — explains the result in plain language
• Download — downloads the result as .xlsx or .csv
• Show Cypher — shows the Cypher query that was run
• Full screen — shows the result in full screen

Continue the conversation

Within a conversation, the assistant remembers what has already been established. You can therefore chain short questions:

Which models are built on the Nova platform?
And on the Atlas platform?
Show it as a chart.

When you change topic, start a New conversation to start from scratch.

@Object

Type @ followed by an object’s name to target a specific object of the graph in your question.

Example:

Which models use the part @ABS-100

@Object type

Type @ followed by an object type to target a specific type of the ontology.

[TODO: difference with @Object]

Example:

List all @Contract negotiated with Bosch, with their annual volume.

/Skill

Type / in the input field to call one of the environment’s skills: pick the skill, then write your question. The assistant then uses that skill, even if the question is phrased differently.

Example:

/cross-brand-part-price-equivalence Which functionally equivalent parts are being purchased at different prices across our brands?

To learn how skills are created and used, see Use a skill.