# Donkye

> Donkye is an AI web-research agent that turns plain-English questions into clean, structured data. It scrapes, crawls, maps, and searches the open web, then returns strict JSON or readable markdown — no code, no selectors, no fragile scripts.

Donkye is built on the Firecrawl v2 extraction engine and Google's Gemini models. The web app handles all browser automation, parallelization, and proxy rotation server-side; API keys never reach the client. Output is designed to be dropped directly into spreadsheets, documents, Google Drive, or downstream AI pipelines.

Typical jobs: lead and account research, competitive monitoring, documentation-to-knowledge-base conversion, market and pricing research, news and changelog tracking.

## Core pages

- [Donkye — home](https://donkye.com/): What Donkye is, how it works, pricing, and product overview.
- [About Donkye](https://donkye.com/about): Company thesis, team, press kit, and investor-facing summary.
- [Use cases](https://donkye.com/use-cases): Worked examples of real web-research jobs Donkye handles.
- [Comparisons](https://donkye.com/compare): Honest comparisons against Firecrawl, ChatGPT browsing, and DIY scrapers.
- [Facts & figures](https://donkye.com/facts): Citable, dated statistics about web research, scraping, and agent tooling.
- [Changelog](https://donkye.com/changelog): Public release history.

## Use cases

- [Lead research](https://donkye.com/use-cases/lead-research): Build target lists with firmographics, tech stack, and contact pages from hundreds of company sites.
- [Competitive monitoring](https://donkye.com/use-cases/competitive-monitoring): Track competitor pricing, product pages, and changelogs without maintaining scrapers.
- [Docs to knowledge base](https://donkye.com/use-cases/docs-to-knowledge-base): Convert any documentation site into a structured JSON knowledge base for RAG.
- [Market research](https://donkye.com/use-cases/market-research): Aggregate prices, reviews, and category data across marketplaces and directories.

## Comparisons

- [Donkye vs Firecrawl (raw)](https://donkye.com/compare/donkye-vs-firecrawl): When to use raw Firecrawl vs the Donkye agent layer.
- [Donkye vs ChatGPT browsing](https://donkye.com/compare/donkye-vs-chatgpt-browsing): Where ChatGPT's browser tool stops and a dedicated agent starts.

## Legal

- [Privacy Policy](https://donkye.com/privacy): What Donkye collects, how it's processed, and how to delete it.
- [Terms of Service](https://donkye.com/terms): Acceptable use, IP, and liability.
- [Security](https://donkye.com/security): Architecture, key handling, and responsible disclosure.

## Machine-readable

- [Full text corpus](https://donkye.com/llms-full.txt): All public Donkye content concatenated as plain markdown for LLM ingestion.
- [Sitemap](https://donkye.com/sitemap.xml)

## Contact

- General: hello@donkye.com
- Privacy: privacy@donkye.com
- Legal: legal@donkye.com
- Security: security@donkye.com