# Vector Search [https://www.elastic.co/what-is/vector-search](https://www.elastic.co/what-is/vector-search) - Vector search powers a more **semantic search**, helping you find the ‘meaning’ of what you’re searching for, so that you don’t have to use the precise search keywords. - The higher the dimensionality of the vector, the more info we have on a document or image and that means the more precise our search will be - As a company who uses enables vector search, you’d need input data, a service that translates that data into a vector or an embedding (using a Transformer model - that’s the meaning of ‘transform’ here) So for example, you may have an input of an ecommerce store with different clothes and such, that would get translated into a document, and then the description would get translated into an embedding. The reason one specific field would become an embedding depends on what actions we want to perform on the embedding. For example, running a filter action on an embedding may not work or be very efficient. In order to set up such a search for an ecommerce store, the first thing we’d need to do is setup our own model, or import one from a repo on HuggingFace or a different provider. After that you can use Elastic’s tooling to import that model into the Elastic cluster. It’s a few minute process and then you have access to the models on the cluster. Step 2: data ingestion and embedding generation after setting up the model you can then choose the data you’re going to be streamlining in and generate the embeddings based on the fields that you point to. we run the docs as they’re coming in through an ‘inference pipeline’ All you have to do is setup a pipeline and select the fields that you want to go through an inference process. Step 3: query the data kNN query is executed in addition to the standard search. Elastic will run the query through the inference pipeline (same as docs), turn them into the same dimensional space so that we can compare with the docs. In addition, the standard filter can be applied so that the results are reduced and performance is optimized. ![Screenshot 2023-01-01 at 18.27.23.png](Vector%20Search%2058f4f6ed0eff47e1b91835c9ad9f6444/Screenshot_2023-01-01_at_18.27.23.png) Hybrid ranking is having vector search alongside the standard search.

‍

‍

Create query function - detailed

  1. Define function
  2. Define logic for candidate filtering
  3. Define logic for candidate scoring
  4. Optional - add aggregation

‍

The rich text element allows you to create and format headings, paragraphs, blockquotes, images, and video all in one place instead of having to add and format them individually. Just double-click and easily create content.

Static and dynamic content editing

‍

How to customize formatting for each rich text

Headings, paragraphs, blockquotes, figures, images, and figure captions can all be styled after a class is added to the rich text element using the "When inside of" nested selector system.