公司：vimeo
Vimeo（/ˈvɪmioʊ/）是一个以视频为中心的社会化网络站点，由IAC/InterActiveCorp拥有并于2004年11月创立。Vimeo支持嵌入，分享，储存视频，也允许用户对视频作出评论。用户必须注册才能上传内容，注册用户也可以创建一个资料页，上传头像，评论，以及“喜欢”一个视频。
Vimeo不允许商业化视频，游戏视频以及色情内容，或者上传任何非用户原创内容。
Behind Viewer Retention Analytics at Scale
Vimeo如何用ClickHouse+AI搞定10亿级视频观众留存分析？关键在于：1. 通过"心跳包"精准捕捉用户行为（跳转/回放/退出），用+1/-1标记观看区间变化；2. 在ClickHouse中用AggregatingMergeTree实时聚合秒级数据，窗口函数计算累计观看量；3. AI环节用滑动平均+RLE压缩数据，结合Prompt工程让Gemini快速生成"反常识"洞察，比如发现第390秒的异常流量高峰。整套方案实现毫秒级查询延迟，1PB数据秒出分析结果。
Unlocking knowledge sharing for videos with RAG
Vimeo’s video Q&A system harnesses the latest AI techniques to enable you to talk to videos about their content.
ClickHouse is in the house
Insights gained and lessons learned from our long video analytics migration journey.
Optimizing object storage
A deep dive into Vimeo’s storage strategy for videos.
A progressive approach to the past
How we moved to a modern transcode system at Vimeo while ensuring cheap and efficient backwards compatibility.
From idea to reality: Elevating our customer support through generative AI
How we prototyped and enhanced the Vimeo Help Desk through rigorous testing.
Riding the dragon
Falkor, Vimeo’s new transcoding infrastructure, maximizes speed while reducing cost.
VMAF FTW
How we turned a metric for video quality into better adaptive streaming.
Enhancing Vimeo Events with layouts
Vimeo Events is a live streaming platform that enables broadcasters to produce and promote virtual events (see Figure 1). From custom registration to post-engagement analytics, Vimeo Events enables any business to build an audience, prepare for an event in advance, and go live easily with a browser-based, studio-quality production tool.
The evolution of event data collection at Vimeo, part 1: the Fatal Attraction era
Our original event tracker racks up a mixed record.
Optimizing for AVX2
How we target CPUs to maximize speed across the software in our stack.
A *bit* about content ratings
A dive into how Vimeo stores and keeps track of a video’s content ratings and respects the visitor’s viewing preference.
dbt development at Vimeo
How we leveled up our data transformation workflow.
One AI, streaming videos for all
Mimir is here to maximize playback quality.
Too big to query: how to query HBase with minimal pain
How to query HBase/Phoenix using Apache Spark from HBase snapshots with minimal impact on the cluster.
Picture-perfect thumbnails for all videos
How we improved thumbnails for HDR, interlaced, and 360° videos.