DOCSEARCHbyAlgolia
Submit Docs

Algolia Documentation

website·Indexed 59 days ago

https://www.algolia.com/doc

Developer guides and API references for building search and AI experiences. Includes documentation on managing data, configuring searchable attributes, custom ranking, faceting, filtering, and implementing geo-search capabilities.

Tokens
1.2M
Snippets
5.1K
Records
7.9K
Agent score
100%

What's inside Algolia

  1. Overview of Algolia Query Suggestions

    Query Suggestions provide a list of popular search terms to users as they type, reducing typing effort and preventing searches that return no results. They are typically implemented in autocomplete menus, but can also be used to provide suggestions before a user starts typing or as a fallback when a search returns no results. Algolia stores these suggestions in a dedicated index that is rebuilt daily based on search analytics, external data, or facet combinations.

  2. Overview of Searcher in InstantSearch Android

    The Searcher interface manages search sessions and processes SearchRequest objects. Algolia provides three primary out-of-the-box implementations: HitsSearcher for searching a single index, FacetSearcher for searching facet values, and MultiSearcher for federated search or query suggestions across multiple indices.

  3. Overview of Algolia Ingestion for Magento 2

    Algolia Ingestion is an optional companion module for the Algolia AI Search & Discovery extension (v3.19+). Instead of writing directly to indices, it routes indexing operations through Algolia's Ingestion Platform. This unlocks three primary capabilities:

    1. Pre-indexing transformations: Use JavaScript in the Algolia dashboard to reshape records before they are indexed.
    2. Up-to-date Collections: Backend hooks for Collections only run on the Ingestion path; direct indexing does not support automatic Collection updates.
    3. Observability: Every indexing operation is logged in the Ingestion runs and events views for inspection.

  4. Overview of Algolia Recommend models

    Algolia Recommend uses supervised machine learning models trained on product data and user interactions to suggest items. It employs two primary algorithm types:

    1. Collaborative Filtering: Analyzes user events (interactions between userToken and objectID) from the last 30 to 90 days to find items that are similar or frequently bought together.
    2. Content-based Filtering: Analyzes key item attributes (like titles or descriptions) to find similarities, which is particularly useful for items with low interaction data (cold start).

    Models are re-trained once every day automatically.

  5. Overview of AI Agent Memory in Agent Studio

    AI agents use memory to retain and recall information across multiple conversations with the same user. This prevents lost context, repetitive interactions, and impersonal experiences. Memory operates in two stages: automatic retrieval at the start of a conversation and active management during the conversation using specific memory tools.

  6. Overview of Algolia A/B Testing

    Algolia A/B testing allows developers to create two alternative search experiences with unique settings, deploy them live, and determine which performs better based on user behavior. It combines relevance tuning with analytics to make data-driven decisions about index configurations, query parameters, and settings.

  7. Overview of Algolia Ask AI

    Ask AI is a conversational AI assistant (chatbot) that enhances keyword search by using an Algolia index to provide context-aware answers grounded in your existing support articles, developer documentation, and blogs. It is LLM-agnostic, allowing you to connect your own LLM provider and API keys to maintain control over costs and behavior. While integrated into DocSearch, it can also be used with custom UIs.

  8. Overview of setting up Algolia Recommend

    Setting up Algolia Recommend involves three primary phases: preparing your data (indexing and event collection), selecting and training the machine learning models in the Algolia dashboard, and integrating the recommendations into your user interface.

  9. Overview of User Authentication in Agent Studio

    User authentication allows Agent Studio to associate conversations and memory with specific users. By providing a secure user token, you can restrict access so users only read their own data and provide personalized experiences in multi-user applications. Without a secure JWT, Agent Studio cannot identify users or scope data per user.

  10. Overview of External Source Groups

    External source groups allow developers to inject specific items into search results based on external logic (e.g., Retail Media Platforms, bidding systems, or ML recommendation engines). Unlike search-based groups, external sources provide precise control over which exact items appear, while Algolia manages the placement, optional re-ranking, and merging with organic results.

  11. Overview of Algolia Merchandising

    Merchandising in Algolia allows you to promote specific products or categories to improve conversion rates, highlight best products, or follow trends. This is achieved through rules that influence the ranking of individual records or entire categories. Rules can be configured in the Algolia dashboard using two methods:

    • Visual Editor: A no-code interface for creating rules and setting search results for specific queries via drag-and-drop.
    • Manual Editor: A more powerful interface where rules are defined as JSON.

  12. Overview of autocomplete-core

    The autocomplete-core package provides the foundational primitives required to build an autocomplete experience. It is primarily intended for developers who are building a custom renderer. For most standard search integrations, using the higher-level Autocomplete library or the experimental Autocomplete widget in InstantSearch (available for both JS and React) is recommended.