Source note
21 days, $5K, 7 AI agents: how a non-programmer built a talent marketplace
Summary
A non-programmer built Bearhug Network, an executive talent marketplace, by managing 7 AI coding agents over a 21-day sprint. The strongest claim is cost and speed: an MVP with more than 75,000 lines of production code for about $5,000.
Problem
- Executive hiring relies on private referrals, search firms, and closed networks, so hiring teams often lack a fast way to browse vetted senior candidates.
- Senior executives also lack access to many roles; the article claims 70-80% of jobs are never posted and more than 90% of executive roles are filled through closed channels.
- Earlier attempts to build this marketplace were too expensive and slow, with estimated costs of a few hundred thousand dollars and 6-12 months of engineering work.
Approach
- The founder used Claude and 7 coding agents as a software production team, while personally directing tasks, testing outputs, and fixing problems.
- The system centralizes Bearhug's recruiting data into an AI-operated internal system called "the brain."
- Bearhug Network shows anonymized, vetted executive profiles that buyers can browse, filter, and request introductions for.
- The Bearhug team screens introduction requests before candidates receive them.
- The build also replaced parts of the firm's sales, marketing, enrichment, and search project management stack.
Results
- Build time: 21 days, with about 350 hours spent managing 7 AI agents.
- Cost: about $5,000 total, compared with the founder's prior estimate of a few hundred thousand dollars.
- Codebase: more than 75,000 lines of production code.
- Data loaded: more than 20,000 executive profiles from 30 combined years of search work by the founders.
- Operations: 6 SaaS tools retired, data enrichment costs reduced by several thousand dollars per year, and overseas admin labor replaced.
- Productivity claim: the founder expects about 80% of administrative time back and at least 2x productive output, but the article does not provide audited metrics or external benchmarks.