Web Search
AI agents with static knowledge quickly become outdated. ChatBotKit's Web Search capabilities give your agents real-time access to information from across the internet, enabling them to answer questions about current events, find recent articles, locate images, and discover videos - all without manual updates to your knowledge base.
When a user asks about breaking news, recent product releases, or anything happening right now, your AI can search the web and return current results. Your agents become genuinely helpful for questions that require fresh information, not just retrieval from pre-loaded datasets.
Key Capabilities
Web Search
Your AI agents can search the general web to find relevant pages, articles, and resources. When users ask questions that require current information - stock prices, recent announcements, current weather conditions - your agent searches the web and incorporates the results into its response. Each result includes the page title, URL, source, and a description to help your agent provide accurate, sourced answers.
News Search
Dedicated news search returns recent articles from news publications. When users ask about current events, industry news, or breaking stories, your agent can search specifically for news content. This surfaces timely journalism rather than general web pages, making it ideal for agents that need to stay current on specific topics or industries.
Image Search
Your agents can find images based on keywords and descriptions. Whether users need product photos, reference images, or visual examples, image search returns relevant results with thumbnails and source URLs. This enables use cases where visual content matters - from finding product imagery to locating reference materials for creative projects.
Video Search
Search for videos across major platforms to find tutorials, demonstrations, or entertainment content. When users need to learn how to do something or want to watch content on a specific topic, your agent can locate relevant videos and share them directly in the conversation.
Dataset Search
Beyond web search, your agents can search your own ChatBotKit datasets using the same search interface. Combine web search for external information with dataset search for internal knowledge, giving your agents access to both current web content and your proprietary information in a unified experience.
Real-World Use Cases
Customer Support with Current Information
A support agent needs to help customers with questions about product availability, shipping times, or recent policy changes. Instead of relying solely on static documentation that may be outdated, the agent searches the company website and recent announcements to provide current answers. Customers get accurate information even when your knowledge base has not been updated.
Research and Analysis Assistant
A research team uses an AI assistant to gather information on market trends, competitor activities, and industry developments. The agent searches news sources for recent articles, finds relevant analysis, and summarizes findings - turning hours of manual research into a conversational query that returns sourced, current results.
Content Discovery Bot
A media company deploys a bot that helps users find relevant content. Users describe what they are looking for - a tutorial on a specific topic, images for a project, or news about a particular subject - and the bot searches across web, image, and video content to surface exactly what they need.
Sales Intelligence Agent
Sales teams use an AI agent to research prospects before calls. The agent searches for recent news about the prospect's company, finds relevant industry articles, and identifies recent developments that could inform the conversation. Sales reps walk into meetings with current, relevant context gathered automatically.
Educational Assistant
An educational platform uses AI to help students research topics. Students ask questions, and the agent searches for recent academic content, news articles, and educational videos that supplement the static curriculum. Learning stays current with real-world developments.
How It Works
Web search is available as abilities you add to your skillsets:
- search/web: General web search for finding pages and articles
- search/news: News-specific search for current articles
- search/images: Image search with thumbnails and sources
- search/videos: Video search across major platforms
When your AI decides a search would help answer a question, it executes the appropriate search ability with the relevant query. Results return as structured data including titles, URLs, sources, and descriptions. Your agent can then use this information to formulate accurate, sourced responses.
ChatBotKit uses Brave Search as the built-in search provider, delivering privacy-respecting results without tracking users. For teams with specific requirements, alternative providers like Serper (Google-based search) are available through the abilities catalogue.
The search abilities work seamlessly with your AI's existing capabilities. Your agent can combine search results with dataset knowledge, memory context, and other abilities to provide comprehensive responses that draw from multiple sources.
Getting Started
Enable web search by adding the search abilities to your skillset. Navigate to your skillset configuration, browse the available abilities, and add the search types your use case requires. Your AI agents will immediately have access to real-time web information.
For most use cases, the built-in Brave Search integration works without additional configuration. If you need Google-based results, you can add the Serper abilities and configure your API credentials.
Consider updating your AI's backstory to describe when and how it should use web search. For example, instruct your agent to search for current information when questions involve recent events, to cite sources when providing search-based answers, or to combine web results with your dataset knowledge for comprehensive responses.
Web search transforms your AI agents from static responders into dynamic research assistants that can access the full breadth of current internet knowledge.