How Exa Indexes the Entire Web: Ray Data and Lance at Billion-Document Scale | Exa | Ray Summit 2026 @anyscale
How Exa Indexes the Entire Web: Ray Data and Lance at Billion-Document Scale | Exa | Ray Summit 2026  @anyscale
Uploaded September 2026 | Updated November 2026, 3 days ago
Maintaining a search index over hundreds of billions of web pages means constantly reprocessing derived artifacts, embeddings, extracted text, and search signals, as content changes. Off-the-shelf data warehouses couldn't meet Exa's need for column-level updates at petabyte scale.

At Ray Summit 2026, Lei Xu, CTO and co-founder of LanceDB, and Hubert Yuan, Software Engineer at Exa, walk through exa-d, Exa's data framework built on Lance and Ray Data: how dependency graphs drive execution, how fragment-level patching avoids write amplification, and how Ray Actors enable stateful model reuse.

You'll leave with patterns for derived-data pipelines at billion-document scale.

Liked this video? Check out other Ray Summit breakout session recordings: youtube.com/playlist?list=PLZNaJyYZRPdo

Subscribe to our YouTube channel to stay up-to-date on the future of AI: youtube.com/c/anyscale

đź”— Connect with us:
LinkedIn: linkedin.com/company/joinanyscale
X: https://x.com/anyscalecompute
Website: anyscale.com
How Exa Indexes the Entire Web: Ray Data and Lance at Billion-Document Scale | Exa | Ray Summit 2026Unlocking Peak Workload Performance & Efficiency with Ray on Kubernetes | Ray Summit 2025Boosting vLLM Inference on Huawei NPU with Ray Compiled Graphs — Huawei | Ray Summit 2025Simulating the Physical World at Scale: Humanoid Locomotion with Ray | Lambda | Ray Summit 2026KV Cache and Long-Context Inference: Solutions for Inference at Scale | VAST Data | Ray Summit 2026Reward Hacking and the Gaps in Agentic AI | Ion Stoica (Anyscale) | Ray Summit 2026How AWS Scales Reinforcement Learning Across Thousands of GPUs | Ray Summit 2025Distributed Model Training with Ray at Capital One | Ray Summit 2025State of vLLM 2026 | Inferact | Ray Summit 2026From Benchmark #1 to Production | Nebius | Ray Summit 2026Foundation Model Distributed Training with Ray | Anyscale | Ray Summit 2026High-Throughput Inference for Synthetic Data & Evals at Sutro | Ray Summit 2025
Anyscale |

How Exa Indexes the Entire Web: Ray Data and Lance at Billion-Document Scale | Exa | Ray Summit 2026

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER