{"id":15531,"date":"2024-09-03T04:30:02","date_gmt":"2024-09-03T11:30:02","guid":{"rendered":"https:\/\/www.couchbase.com\/blog\/?p=15531"},"modified":"2025-06-13T22:41:59","modified_gmt":"2025-06-14T05:41:59","slug":"vector-search-at-the-edge-with-couchbase-mobile","status":"publish","type":"post","link":"https:\/\/www.couchbase.com\/blog\/vector-search-at-the-edge-with-couchbase-mobile\/","title":{"rendered":"Vector Search at the Edge with Couchbase Mobile"},"content":{"rendered":"

We\u2019re pleased to announce the <\/span>release of Couchbase Lite 3.2<\/span><\/a> with support for vector search. This launch follows the coattails of <\/span>vector search support on Capella and Couchbase Server 7.6<\/span><\/a>.\u00a0 <\/span>Now, with vector search support in Couchbase Lite, we enable cloud to edge support for vector search powering AI applications in the cloud and at the edge<\/b><\/span>.<\/b><\/p>\n

In this blog post, I will discuss the key benefits of supporting vector search at the edge, including a brief look at use cases that fall within your Couchbase Lite applications.<\/p>\n

What is Vector Search?<\/span><\/h3>\n

Vector search<\/span><\/a> is a technique to retrieve semantically <\/span>similar items based on <\/span>vector embedding<\/span><\/a> representations of the items in a multi-dimensional space. Distance metrics are used to determine the similarity between items. Vector Search is an essential component of <\/span>Generative AI<\/span><\/a> and Predictive AI applications.\u00a0\u00a0<\/span><\/p>\n

Couchbase Mobile Stack<\/span><\/h2>\n

If you are new to Couchbase, here is a quick primer on Couchbase Mobile.\u00a0<\/span><\/p>\n

\"\"<\/a><\/p>\n

Couchbase Mobile is an <\/span>offline-first<\/span><\/a>, cloud-to-edge database platform. It is comprised of the following:\u00a0<\/span><\/p>\n

Cloud Database<\/b>: Available as a fully managed and hosted Database-as-a-Service with <\/span>Couchbase Capella<\/span><\/a>, <\/span>or deploy and host<\/span> Couchbase Server<\/span><\/a> on your own.<\/span><\/p>\n

Embedded Database:<\/b> Couchbase Lite<\/span><\/a> \u00a0is a full featured, NoSQL embedded database for mobile, desktop and IoT applications.\u00a0<\/span><\/p>\n

Data Sync:<\/b> A secure gateway for data sync over the web, as well as peer-to-peer sync between devices. Offered as fully hosted and managed sync with<\/span> Capella App Services<\/span><\/a>, <\/span>or install and manage<\/span> Couchbase Sync Gateway<\/span><\/a> yourself.<\/span><\/p>\n

Check out our <\/span>documentation<\/span><\/a> for more information.<\/span><\/p>\n

Vector Search Use Cases & Benefits<\/span><\/h2>\n

While the benefits of <\/span>vector search<\/span><\/a> are fairly well understood, why would you want vector search at the edge?\u00a0<\/span><\/p>\n

Semantic Search in Offline-First Mode<\/span><\/h3>\n

Applications where simple text-based searches are insufficient can now support semantic searches on local data to retrieve contextually relevant data even when the device is in offline mode. This ensures that the search results are always available.\u00a0<\/span><\/p>\n

Example<\/h4>\n

Consider a classic field application\u00a0 Utility workers out at repair sites and disaster areas operate in areas with poor or no Internet connectivity:<\/span><\/p>\n