CREATION PIPELINE

Sylva AI’s content creation pipeline follows a hybrid architecture, combining secure on-chain prompt registration with efficient off-chain AI processing. The pipeline consists of two primary phases: Prompt Submission and Response Retrieval.

1. Prompt Submission

Users initiate a creative request (text, image, or audio) through the User Contract by invoking the writePrompt() function. This request is forwarded to the On-Chain Service Contract, where:

• Required fees are deducted, and

• Prompt metadata (e.g., type, hash, timestamp) is securely recorded.

The service contract then emits an event, which serves as a trigger for the off-chain infrastructure. A backend server continuously listens for these events, retrieves the full prompt, and routes it to the appropriate AI processing engine for generation.

2. Response Retrieval

Once the content is generated and indexed, users call the readResponse(promptId) function via the User Contract. The system returns the generated output (e.g., audio, video, or narrative) corresponding to the original prompt ID, provided the transaction is complete and the output is available.