Enabling Multimodal Search for a Stock Media Platform with Viking AI Search--Viking AI Search-Volcengine

Docs

Docs Console

Viking AI Search

Document

Viking AI Search

Document

Viking AI Search Documentation

Product Introduction

Viking AI Search Product Introduction

Pricing

Viking AI Search (Standard Edition)

Viking AI Search (Free Trial)

Upgrade to Standard Edition

Quick Start Guide

Configure Personalised Recommendation Capabilities

Configure Multimodal Search Capabilities

Configure Conversational Q&A Capabilities

Create an AI Search Application

User Guide

Prepare Datasets

Overview of Item Datasets

Introduction to Image & Text Item Dataset

Introduction to Video Dataset

Introduction to Behavior Dataset

Create an Application and Link Item Datasets

Configure AI Search

Introduction to Viking AI Search

About datasets and indexes

Configure search strategy

Search Experience Page Settings

Integrate AI Search

AI Search: Filter

Search: Search Results Sorting

Search: Boost and Demotion

Configure Personalized AI Recommendation

Introduction to AI Recommendation

Prerequisites and setup

Experience Centre

Features Overview

Managing Recommendation Scenarios

Recommendation Input

Retrieval Strategy

Configure Fallback Strategy

Item Filtering

Configure Static Filter Rules

Configure Dynamic Filter Conditions

Recommendation Deduplication

Remove items already exposed to users

Item Re-ranking (Reordering)

Pinned Recommended Items

Item Attribute Boost and Demotion

User geographic location-based ranking boost and demotion

Time-based Boost and Demotion

Recommendation API uplift and demotion rule configuration

Recommendation diversity

Diversity rule configuration

Configure Conversational Search (Agentic Search)

Introduction to Conversational Search Assistant (Agentic Search)

Configure and experience the conversation assistant

Configure Conversational Search Assistant instructions

Configure Conversation Opening

(Beta) Use Documents as Knowledge Base

Integrated Conversational Search

Dataset

Dataset Introduction

Create Item Dataset - Image & Text Data

Create Item Dataset - Video Data

Import Item Data

Import JSONL Files using Console

Import and Update Item Data in Real Time

Batch Import Item Data

Document Dataset

Create User Behavior Dataset

Import User Behavior Data

Import Behavior Data JSONL Files using Console

Import Behavior Data using API

Synthesize New Fields using Large Language Models

App Management

Monitoring

Best practices

Building an E-commerce agent with Viking AI Search

Integrating Personalized recommendations in social applications using Viking AI Search

Enabling Multimodal Search for a Stock Media Platform with Viking AI Search

Create an AI companion for a video platform using Viking AI Search

Configure and Integrate Conversational Search

API reference

General Introduction

API Authentication

Common Error Codes

Data API

Write - Import and update data in real time

Delete - Delete specified data

ListItems - Get the list of items in a dataset

GetItem - Get Item Details

ListDatasetData - Get video data and details (Not recommended)

Batch data import

CreateBatchImport - Create Batch Import Task

BatchImport - Batch data import

CompleteBatchImportTask - Complete Batch Import

GetBatchImportStatus - Check Import Status

OpenAPI

BrowseIndex - Browse item data

Troubleshooting guide for common issues

Meaning and handling of common error messages

About datasets

Multimodal search scenarios

Personalized recommendation scenarios

Conversational search scenarios

Why did the recommendation not return any items?

Documentation

Viking AI Search

Viking AI Search Documentation

Best practices

Enabling Multimodal Search for a Stock Media Platform with Viking AI Search

Copy page

Download PDF

Best practices

Enabling Multimodal Search for a Stock Media Platform with Viking AI Search

Copy page

Download PDF

Enabling Multimodal Search for a Stock Media Platform with Viking AI Search

Last updated: 2026.04.13 22:17:33

Public security network record in Beijing: No. 11010802032137