This example demonstrates how to build an Upsonic LLM Agent that autonomously finds and validates a company's official website using reasoning and lightweight search tools.

The agent:

• Searches for potential websites using the Serper API.
• Reasons through results to identify the most credible domain.
• Validates the site based on brand–domain matching and context.
• Returns a structured, explainable JSON output.

uv sync

Copy the example environment file and edit it:

cp .env.example .env

Then open .env and replace the placeholder with your real API key:

SERPER_API_KEY=your_api_key_here

You can get a free key at https://serper.dev.

Run the reasoning-based agent to find a company's official website:

uv run task_examples/find_company_website/find_company_website.py --company "OpenAI"

Example Output:

{
  "company": "OpenAI",
  "website": "https://openai.com/",
  "validated": true,
  "reasoning": "The domain 'openai.com' directly matches the company name 'OpenAI', indicating a strong likelihood that it is the official website.",
  "confidence": 0.95
}

• Tool Layer – A single function (get_company_candidates) queries Serper for candidate URLs.
• Agent Layer – The Upsonic agent performs reasoning, filtering out irrelevant sites and selecting the most official one.
• Schema Layer – Results are returned in a structured WebsiteResponse format (company, website, reasoning, confidence).

This structure demonstrates how Upsonic agents can mix retrieval, reasoning, and structured outputs in one clean workflow.

task_examples/find_company_website/
├── find_company_website.py       # Legacy finder (modular version)
└── README.md                     # This file

# Root directory
.env.example                      # Example environment config

• The demo emphasizes agent reasoning, not manual rule-based filtering.
• You can easily extend the agent by adding tools (e.g., WHOIS checks, HTML analyzers).
• Ideal for showing how LLMs can autonomously use tools and justify their decisions.