The AI-Powered Lucid Conversion feature provides a faster and more efficient method for creating IVR flows by allowing users to convert externally designed flow diagrams into a structured IVR within the platform.

Traditionally, IVR creation requires users to manually construct the entire call flow inside the IVR Builder. This involves adding nodes one at a time and configuring prompts, conditions, and routing logic for each step. While this approach offers full flexibility, building large or complex IVRs can be time-consuming and increases the possibility of configuration inconsistencies.

With AI-Powered Lucid Conversion, users can design the IVR flow using external diagramming tools such as Lucidchart or any tool capable of exporting an XLSX file. The diagram is then exported as a CSV file and uploaded to the platform. The system parses the uploaded file, interprets the node definitions and relationships, and automatically generates a corresponding IVR structure inside the IVR Builder.

From a technical perspective, this feature is powered by a Retrieval-Augmented Generation (RAG) architecture integrated with Large Language Models (LLMs). The uploaded CSV data is analyzed to identify node types, connections, and flow dependencies. Based on this interpretation, the system constructs an initial IVR structure and automatically creates the required Dialogue Tree records in the backend. These records are then rendered as nodes and connections within the IVR Builder UI.

The generated IVR acts as a draft configuration. Users are expected to review the generated flow, validate node configurations, and refine prompts or routing logic based on their specific business requirements before deploying the IVR.

By automating the initial flow generation, the feature helps reduce manual setup effort, improves structural consistency, and accelerates IVR implementation. This allows users and implementation teams to focus on refining business logic and optimizing the customer interaction experience rather than rebuilding flow diagrams within the platform.

Creating IVR through AI-Powered Lucid Conversion

1. Create the IVR flow diagram in an external diagramming tool (such as Lucidchart) by defining the required nodes and connecting them to represent the call flow logic.

2. Export the completed workflow from the diagramming tool in CSV format so it can be uploaded and processed by the platform.

   Note: Before exporting the workflow, verify that all nodes in the diagram are correctly defined and properly connected. This ensures the system can accurately interpret the flow and generate a more reliable GenAI-suggested IVR structure.

   

3. Follow the steps mentioned in the Create New IVR section to begin setting up a new IVR configuration.

4. On the IVR Details screen, click Import from Lucid and upload the CSV file.

   

   ​

   

5. The system parses the CSV and:

   1. Identifies nodes.

   2. Maps relationships.

   3. Infers IVR node types.

   4. Builds a draft dialogue tree.

6. The system processes the uploaded file and begins importing the flow, automatically generating the corresponding IVR structure.

   

7. After the import process is completed, the IVR status appears as ready for review, indicating that the generated flow is available for validation and further configuration.

   ​

8. After the IVR is generated:

   1. Open each node individually.

   2. Review the configuration.

   3. Verify prompts, conditions, and routing.

   4. Make necessary corrections.

9. The IVR cannot be deployed until every node has been reviewed. Once all nodes are checked and verified, deployment becomes available, ensuring that the AI-generated structure is validated before going live.

   

10. After all nodes have been reviewed and validated, the Deploy option is activated. You can then save the IVR and publish it for live use.