Agentic Internet Rsearch

Agentic Internet rsearch is an intelligent deep agents web search and analysis system that combines real-time retrieval with autonomous agents to deliver comprehensive research insights. The system automatically searches multiple sources, analyzes content using advanced AI models, and presents structured findings with citations and save all the work and calculate the metrics to evaluate the full search process.

Key Features:

• Real-time web search using SerpApi for current information
• Search and Data indexing with Algolia
• Autonomous agents that can follow research threads and cross-reference sources
• Automated fact-checking and source verification
• Interactive dashboard for exploring research results in opensearch by Using Mistral Le chat as a MCP Client

Technologies Used:

• SerpApi for real-time search results
• Algolia for data indexing and semantic search
• Google AI models for content analysis
• Scalingo for opensearch metrics and index.
• Python Langchain
• Mistral Le Chat for monitoring the OpenSearch MCP Server with IA.

Demo: The system can research complex topics in under 2 minutes, providing structured reports with verified sources. Judges can test by entering any research query and observing the autonomous research process.

Code Repository: https://github.com/team/agentic-internet-search
Performance: Average query response time <30 seconds, 95% source accuracy rate.