James Murray introduces an AI-driven system that generates narrated poetry, music prompts, and visual themes to support creative expression. Perfect for artists, musicians, and writers looking to enhance their creative processes with AI assistance.

The system uses multimodal AI to generate synchronized outputs: a poem, its spoken narration with emotional prosody, a music prompt in ABC notation or MIDI, and a visual mood board generated via Stable Diffusion. All outputs are time-aligned for seamless performance or recording.

Users input a theme, emotion, or memory -- the AI returns a complete creative package ready for recording, performance, or social sharing. The system supports 12 poetic forms and 8 musical genres.

Key Features

• Poetry Narration: ElevenLabs V3 with SSML emotional tags for natural cadence.
• Music Prompt Generation: Outputs ABC notation, MIDI, or chord progressions with tempo and key.
• Visual Theme Creation: Stable Diffusion XL generates 512x512 mood boards.
• Emotional Prosody Engine: Adjusts pitch, pace, and pauses based on sentiment analysis.
• Style Transfer: Apply "Rumi", "Beat Poetry", "Baroque", or "Trap" across modalities.
• Export Bundle: Download .mp3, .mid, .png, .txt, and .pdf in one click.

System Design & Architecture

The system integrates multiple AI models via a central orchestrator. A prompt is expanded into a structured JSON spec, then dispatched to specialized models. Outputs are synchronized using a timeline engine.

Technical Stack

• Text: GPT-4 + fine-tuned on 50k poems
• Voice: ElevenLabs V3 + SSML
• Music: MusicGen + custom genre classifier
• Visual: SDXL + ControlNet for composition

Related Projects

Explore other related projects:

• AI Voice Production
• Video Production
• ElevenLabs Narration