Backend Developer for ComfyUI JSON Workflow Integration

Type the result of 2+4 on top of your application to make sure that you are not a bot
Looking for a developer who can integrate JSON workflows into our system and enable automatic LoRA (.safetensors) selection based on the prompt.

The goal is to have a backend endpoint that accepts a user’s text prompt, analyzes it, updates the JSON flow dynamically (replacing model nodes, prompt text, and LoRA loaders), and sends the modified flow to the API for rendering. The backend should then return the generated video result.

You should have experience with:

ComfyUI or Stable Diffusion pipelines.

Modifying and managing JSON flows programmatically.

Handling LoRA (.safetensors) model files dynamically.

Implementing basic prompt understanding logic (keyword-based or simple NLP) to select the correct LoRAs automatically.

Building Python or Node.js backends that communicate with GPU endpoints (e.g., hosted GPU runtime like RunPod, Brev, or Replicate).

The ideal candidate can make the workflow dynamic, efficient, and ready for production, ensuring that the backend can select the right LoRAs and modify JSON structures in real-time before sending to ComfyUI.

Paying $100