icpfinder

icpfinder is a free, open-source command-line and web tool aimed at independent software builders who want to find their first paying customers before investing heavily in marketing. You paste a one-line description of your product idea, and the tool returns three types of potential customers (called archetypes) along with verified email addresses for real people at matching companies, all within about 30 seconds.

Internally, the tool sends your product description to an AI model (Google Gemini by default) which identifies three distinct customer profiles, each with an industry, a job role, a pain point your product might address, and signals to look for when prospecting. For each profile it then queries an email verification service (Hunter.io by default) to find actual decision-makers and confirm their contact addresses. The whole process streams results back as they arrive rather than waiting until everything is finished.

The codebase is split into three packages. The core package contains the orchestration logic and defines the interfaces any AI or email provider must implement. The providers package contains the default adapters for Gemini and Hunter.io. The web package is a Next.js application that exposes the tool as a hosted demo, with rate limits per IP address and optional caching to avoid repeat API calls for the same seed text.

Using your own API keys for Gemini and Hunter.io costs roughly a dollar per run, mostly from the email verification lookups. A stub mode lets you run the demo locally for free without any keys, though the results will be placeholder data. The hosted demo is also free but caps usage per IP.

The project is MIT-licensed and written in TypeScript. It is designed so that the underlying AI and email providers can be swapped without modifying the core engine, which makes it straightforward to substitute other services. With 17 stars on GitHub, it is a recent release at an early stage.