VectraFlow: Long-Horizon Semantic Processing over Data and Event Streams with LLMs
Shu Chen, Junhan Liu, Deepti Raghavan, Ugur Cetintemel · 2026
Read VectraFlow: Long-Horizon Semantic Processing over Data and Event Streams with LLMs by Shu Chen, Junhan Liu, Deepti Raghavan, Ugur Cetintemel (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-03855.txt.
Description
Shu Chen and colleagues demonstrated VectraFlow, a streaming engine that watches text streams over long stretches and remembers what it has seen. Today's language-model frameworks answer one prompt at a time with no memory, while classic event systems demand neat typed events. VectraFlow extends familiar database operators with language-model execution, offering filter, map, aggregate, join, group-by, and window steps. A semantic event pattern operator adds time-based rule matching over events pulled from text. The team showed it live on clinical documents.
Catalog record
| CREATOR | Shu Chen, Junhan Liu, Deepti Raghavan, Ugur Cetintemel |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 16594 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.03855