How to Build Agentic Web Research Assistants | Live Demo & Code

This practical session from the AI Agents Conference 2025 demonstrates how to build agentic research assistants capable of searching the web, reasoning over multiple sources, and delivering structured, reliable results.

In this hands-on talk, Frank Boucher walks through a real-world demo of a web-based research agent that helps users discover upcoming tech events using live web data, reasoning steps, and clean architectural patterns.

The session focuses on how agentic systems differ from traditional LLM calls by performing iterative searches, aggregating results, and explaining their reasoning—making them ideal for research-heavy applications.

🔹 Agenda Covered in This Session

Live demo of an Event Finder research agent
How agentic research assistants search the web iteratively
Understanding the platform and models powering the demo
Handling user location, domains, and filtering logic
Structured JSON responses for production-ready apps
Reasoning visibility for debugging and trust

🔹 What Makes Agentic Research Assistants Powerful

Multi-step web research instead of single LLM calls
Context-aware search using location and time filters
Domain allowlists and blocklists for precise retrieval
Grounded answers with URLs and source attribution
Adjustable depth of research for speed vs accuracy

🔹 Technical Deep Dive Highlights

API-based agent design compatible with OpenAI-style formats
Web search agents without full SDK dependency
JSON schema enforcement for predictable outputs
Integration with Blazor and modern .NET applications
Logging and reasoning inspection for debugging

🔹 Who Should Watch This Session

Developers building AI-powered research tools
Engineers working on agentic workflows and assistants
Architects designing AI-first web applications
Teams exploring grounded AI and hallucination reduction
Anyone interested in real-world AI agent demos

This session is ideal for developers, AI engineers, cloud architects, and engineering leaders looking to build production-ready agentic research systems that go beyond simple chatbots.

🎙 Speakers

Frank Boucher
Developer Advocate | Cloud Solutions Architect, Reka

📌 Event

AI Agents Conference 2025
Powered by C# Corner

LIKE | COMMENT | SHARE | SUBSCRIBE!!!

► CSharp Corner: https://www.c-sharpcorner.com
► Facebook: https://www.facebook.com/CsharpCorner
► Twitter: https://x.com/CsharpCorner
► LinkedIn: https://www.linkedin.com/company/csharpcorner
► Instagram: https://www.instagram.com/csharpcorner