This use case shows how to build a small price-intelligence app that searches eBay, extracts marketplace signals, and turns raw listing results into a structured summary.
Use it when you need to monitor marketplace prices, compare listings, track competitors, or give an AI agent fresh ecommerce context without maintaining your own scraper.
Live example
- Live demo: ebay-example.desearch.ai
- Example repository: Desearch-ai/desearch-ebay-price-intelligence-example
- APIs used: Web Search and optional Web Crawl
What the example does
The demo turns a product query into a marketplace research workflow:
- The user enters a product query, such as an iPhone, GPU, sneaker, or collectible.
- The app searches eBay-focused web results with Desearch Web Search.
- It extracts useful listing signals such as title, URL, condition, price-like values, and source snippets.
- It summarizes the result set with basic price intelligence, including lowest, highest, average, and median-style signals where available.
- It presents the results as cards and summary metrics so the user can quickly compare the market.
Why this is useful
Marketplace data changes constantly. Product teams, ecommerce operators, and research agents often need fresh listing context before making decisions.
With Desearch, you can build workflows such as:
- Price monitoring — check how a product is currently priced across marketplace listings.
- Competitive research — compare similar products, sellers, and listing angles.
- Inventory analysis — identify common price ranges and conditions for used or refurbished items.
- Agent workflows — give an AI agent current marketplace context before it recommends a buy, sell, or research action.
- Market snapshots — generate quick summaries for product categories without building a crawler from scratch.
Architecture
Browser UI
-> Example app API route
-> Desearch Web Search API
-> Optional Web Crawl for selected URLs
-> Normalized price-intelligence summary
-> Results rendered in the demo UI
The example keeps the Desearch API key on the server side. The browser talks to the example app, and the app calls Desearch from a backend route.
API pattern
The example starts with Web Search because eBay discovery is a search problem first. Use a focused query such as:
site:ebay.com Apple iPhone 15 Pro Max
A backend route can call Desearch, normalize the returned links/snippets, and optionally crawl selected result URLs when deeper page text is needed.
curl --request GET 'https://api.desearch.ai/web?query=site%3Aebay.com%20Apple%20iPhone%2015%20Pro%20Max&count=10' \
--header 'Authorization: $DESEARCH_API_KEY'
Keep API keys server-side. Do not expose production keys in browser JavaScript or public repositories.
What to build next
You can adapt this example into:
- a price tracker for a fixed product list,
- a competitor monitoring dashboard,
- an ecommerce research assistant,
- a product sourcing workflow,
- or an agent tool that retrieves fresh marketplace context on demand.
Start from the repository, replace the sample query presets with your own product categories, and add storage or alerts if you need ongoing monitoring.