QCraft and MaxCompute Introduce Embodied AI UMI Data Pipeline Practice
11/17/2026 — 11/17, 17:17·1 sources (all)·1 reports (all)
Key takeaways
- At the 2026 Apsara Conference Big Data Forum, QCraft AI scientist Lyu Jun presented a practice using MaxCompute to batch-process videos and…
Background & analysis
On November 17, 2026, a report published on Juejin covered a presentation delivered at the big data forum of the 2026 Apsara Conference. Lu Jun, an AI scientist at Qingche Intelligence, used the UMI data pipeline as his example to explain how the team processes embodied AI UMI data at scale on MaxCompute. According to the presentation, the system batch-processes videos and action trajectories from the UMI data pipeline on MaxCompute, and leverages full-modal computing to turn real-world operation recordings into high-efficiency datasets suitable for model training. The available reporting covers only this forum presentation, and there is no verifiable information yet on how the practice will be rolled out or advanced going forward.
AI-generated from 1 reports · updated 13 hours ago
Related AI tools
- Heat index
- 66
- All sources
- 1
- All reports
- 1
- First seen
- 14 hours ago
Reports on this story headlines open the original
At the 2026 Apsara Conference Big Data Forum, Pintia AI scientist Lü Jun presented how the team processes video and action trajectories for the embodied AI UMI data pipeline using MaxCompute. The practice converts real-world operation records into structured datasets for model training through multimodal computing.
掘金AI score 55
Other stories people are talking about
- Heat index 65Musk Says Tesla May Pilot Private Car Access to Robotaxi Next Month1 sources
- Heat index 1331Nvidia board establishes national security and export control committee25 sources
- Heat index 78Hugging Face Releases Pi Agent LoRA Adapter for Qwen3.8-27B1 sources
- Heat index 74使用 Laravel AI SDK 构建大规模新闻处理管道1 sources
- Heat index 68A Deep Dive into Context Engineering for AI Agents1 sources
How is heat calculated?About the methodHide
Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.
This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.
- Surge
- Discussion rising fast
- New
- First report within 6 hours
- Rising
- Still gathering discussion

