A reference guide to the common features, tabs, and behaviors of workflow nodes in Rayven.
Overview
Workflow nodes are the building blocks of all application logic in Rayven. Each node performs a defined role—whether that’s connecting to an external system, applying logic, embedding AI, or displaying data in an interface. Nodes pass information between each other as JSON payloads, and together they form workflows that drive both backend automations and frontend interfaces.
While every node has unique capabilities, there are a set of features, tabs, and behaviors that are consistent across all nodes. This guide explains those commonalities so you can understand how to configure and work with any node in the Rayven platform.
Node Categories
Rayven’s workflow nodes fall into four categories:
- Connector Nodes – handle data ingestion or output (APIs, FTP, HTTP, MQTT, etc.).
- Logic Nodes – transform, calculate, filter, or manipulate payloads.
- AI Nodes – interact with external AI services (e.g. OpenAI).
- Frontend Nodes – create widgets and interactive elements for your application interfaces.
Common Node Tabs
Every workflow node in Rayven has three standard tabs:
1. Setup
- The configuration area for the node.
- Contains all adjustable parameters, such as connection details, query logic, formatting, or calculation methods.
- Each node type has its own unique setup options, but the tab structure is consistent across the platform.
2. Inspect Data
- Lets you view the payloads flowing in and out of the node in real time.
- Displays recent incoming and outgoing JSON payloads, with filtering options (e.g. by UID, date range, or key text).
- Useful for debugging workflows
3. Help
The Help tab provides reference details for the node and quick access to documentation. It contains:
- Workflow ID – the unique identifier for the workflow this node belongs to.
- Node ID – the unique identifier for this node, useful for referencing when exporting workflow data or sending data between workflows.
- Knowledge Base Link – a direct link to the user guide article for the node, giving you configuration instructions and best practices.
Data Flow Through Nodes
Rayven workflows move data between nodes as JSON payloads. Understanding how nodes reference and manipulate these payloads is key to building correct logic.
JSON Key Referencing in Nodes
- Each value in the payload is stored under a JSON key.
- Many nodes include fields such as:
- “JSON Key” – tells the node which input value to use.
- “Output JSON Key” – defines the name under which the node will store its result.
- “JSON Key to Display” – common in Frontend nodes to determine which value appears in the widget.
- As data flows through the workflow, new values are usually appended to the payload under their specified output key, rather than overwriting existing values
Dynamic Referencing with Special Syntax
Some nodes support dynamic referencing of keys from the incoming payload, allowing you to control node behavior with data instead of static values. Reference the user guide article for detail:
Node Syntax: JSON Key References and Dynamic Expressions
Accumulation of Data
- As payloads move downstream, they accumulate values from previous nodes.
- This lets you build rich, structured payloads that combine data from multiple sources before sending them onward or storing them in a table.
Storage & Access
- All workflow data is automatically stored in the Cassandra time-series database, indexed by:
- UID (Primary Table anchor)
- Node ID (the node where it was processed)
- Timestamp (the time it was processed)
- This makes every node’s outputs retrievable over time, either for historical analysis or for populating frontend widgets.
Iterative Execution
- Many nodes (especially Trigger and Connector nodes) can be configured to run:
- Once per interval
- Once per Primary Table row (UID), per interval
- Once per subset of rows matching Label Values, per interval
- This allows workflows to be applied at the right level of granularity—either globally, per entity, or per group.
Best Practices for Working with Nodes
- Name nodes clearly – helps when inspecting data and debugging.
- Use Inspect Data often – to validate payloads and confirm logic is working as expected.
- Start simple – build small chains of nodes, confirm their behavior, then expand.
- Be mindful of iteration – understand whether your node will run once, per UID, or per Label Value.
- Check JSON key syntax – always use the right format (quoted, unquoted, or dynamic) for the node you’re configuring.
Summary
All Rayven workflow nodes share a common structure, making them predictable and easy to work with once you understand the basics. Regardless of whether you’re configuring a connector, applying logic, calling AI, or creating a frontend widget, every node provides:
- A Setup tab for configuration
- An Inspect Data tab for debugging and validation
- A Help tab with IDs and documentation links
Data flows between nodes as JSON payloads, with keys referenced explicitly in node fields. By using the correct syntax and understanding how payloads are built up across a workflow, you can reliably configure even complex automations and interfaces.
FAQs
Q: Do all nodes support both inputs and outputs?
A: No. Nodes are visually designed to show what they support:
- A left connector means the node accepts incoming payloads (input).
- A right connector means the node produces an outgoing payload (output).
- Some nodes have both, meaning they can receive data, process it, and then pass it on.
- Others may only have one side (e.g. Trigger nodes have only an output, since they start workflows).
So by looking at the node shape in the workflow builder, you can immediately see whether it handles inputs, outputs, or both.
Q: Where can I find a node’s Workflow ID and Node ID?
A: Inside the Help tab of the node.
Q: How do I debug a node?
A: Use the Inspect Data tab within the node. It shows recent payloads and lets you filter by UID, date, or text.
Q: Is data automatically stored in every node?
A: Yes. Once the node has been added to a workflow and starts receiving payloads, its data is stored automatically.
Q: Do I need to configure storage separately for a node?
A: No. Storage in Cassandra is automatic. To store results in structured Tables, use the Update Table nodes.
A: In general, the standard input and output JSON key fields in nodes do not require any special syntax — you simply type the JSON key name directly. However, some nodes include additional dynamic fields (for example in query or filter settings) that do require special syntax. The exact rules depend on the node. Refer to the article “Node Syntax: JSON Key References and Dynamic Expressions” for the specific syntax requirements of each node.