Chapter 5 — BlockFlow · Node Reference
5.2.2 Input nodes
Input nodes receive information from users and form the foundation of interactivity. They enable user interaction, dynamic flow control based on responses, structured or unstructured data collection, personalisation, and decision-making when paired with logic nodes.
The Input node in the BlockFlow designer.
Input Text node
Receives and processes text-based input. It prompts the user for specific information or waits for an open-ended response, validates the input against criteria such as a format, keyword or regular expression, and stores the response in a variable for later nodes.
Input Query node
The scenario’s decision-maker: it interprets input, identifies intent and routes the conversation.
- Intent recognition — NLP maps the user’s intent to predefined intents in the flow.
- Dynamic redirection — the conversation continues at the node connected to the identified intent.
- Fallback handling — unmatched input is routed to a fallback node or answered generically.
- Context awareness — context is maintained for accurate recognition across multi-turn conversations.
- Customisable intents — intents are defined and mapped to actions or nodes.
Input Choice node
Captures a selection from a set of dynamic options — useful when the choices vary by context or input data.
- Dynamic option generation — options built from context or prior input, such as a list of products or services.
- Intent recognition for options — recognises varied phrasing; for “Yes” and “No” it also understands “Yeah” or “Nope”.
- Flow redirection and fallback handling — routes on the selected option, or asks for clarification when nothing matches.
- Multi-language support — selections can be understood in different languages when configured.