DataMagic: Authoring Data Videos through Declarative Multi-Agent Orchestration
DataMagic authors accurate data videos from tables via declarative multi-agent orchestration, beating GPT-5 baselines.
DataMagic authors data videos from raw tabular data using a declarative specification, DVSpec, and multi-agent orchestration. DVSpec jointly represents charts, narration, and animations with data-bound references so provenance and audio-visual alignment are preserved. A generate-then-orchestrate strategy produces candidate scenes in parallel and then optimizes narrative coherence, with shared state for automated and human-controlled editing. On 109 real samples, quality rises from 2.13/5 for advanced LLMs such as GPT-5 to 3.89/5, execution success exceeds 95%, and a user study finds a 79.7% reduction in task time.
- DataMagic turns raw tables into narrated, animated data videos.
- DVSpec binds charts, narration, and timing for provenance and alignment.
- Generate-then-Orchestrate agents draft scenes, then optimize narrative coherence.
- Quality rises to 3.89/5 and success exceeds 95% versus GPT-5 baselines.
- User study reports 79.7% less task time and lower cognitive load.
From the original article
1 min read at huggingface.co
Data videos communicate data insights through dynamic charts, voice narration, and synchronized animations, and have become a widely adopted form of data storytelling. However, producing them requires expertise in data analysis, narrative design, and video editing.
By Yupeng Xie, Zhenyang Wang, Liangwei Wang · huggingface.co
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.