An experiment: an AI job agent that reads every advert before applying
Hiring has turned into a numbers game. UK graduate roles draw a record 140 applications each, up to 290 in the busiest sectors, and in the UK test run described below employers regularly mentioned 300 or more. Many candidates answer with bots that apply to thousands of jobs overnight; the published results are around one interview per 250 applications, or worse. Watching friends go through this,…
Titled "An experiment: an AI job agent that reads every advert before applying," a new open-source tool called JobAgent has been created to streamline the job search process. The tool, built on Claude Code, analyzes job advertisements in full, scores the fit, and writes a tailored CV before proceeding with applications. This approach aims to reduce spam and increase interview opportunities.
The experiment was inspired by Santiago Fernández and his career-ops system, and JobAgent follows a similar philosophy but is designed to work with Claude Cowork, a desktop app for Claude. Users interact with the tool by talking to Claude, who handles all the technical aspects, including setup and applying to roles.
In a UK test run, four individuals used JobAgent on their own job searches over 13 weeks. The results show that out of 10,483 roles that passed initial filters, 8,585 job advertisements were read and scored as a fit. Of these, 779 roles were tailored with a customized CV, and 258 applications were submitted by hand. This leads to roughly 31 interviews per 1,000 applications, compared to the typically 1 to 4 interviews per 1,000 applications for overnight bots.
The tool emphasizes reading job advertisements before applying, which appears to be the key to its success. Other testers using JobAgent also saw similar results, with interview-to-application ratios ranging from 0.5% to 12%, depending on the individual and the market. JobAgent is currently in the testing phase, with no benchmark data, but it offers a semi-autonomous job-application pipeline that requires no coding skills.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
