hireburn

What it does

hireburn is a dependency-free Python toolkit (Python 3.10+, standard library only) for AI-assisted hiring. It reads CSV, YAML or JSON and writes Markdown and JSON. It makes no network calls, so nothing leaves the machine. MIT licence.

The method comes from a published case report in which the labels of an AI interview did not line up with who was still working weeks later. hireburn is part of the payrollburn case series. Example data in the repository are synthetic.

Three parts

avatar is a record-once kit for AI-avatar first interviews. It has a recording script and checklist, an interview plan of pre-recorded clips with at most one voice-AI follow-up per question, a candidate page with mandatory AI disclosure and a consent gate, and a transcript rubric. The HIRE / MAYBE / NO labels carry the caveat that a label is not a measurement.

outreach prepares personal LinkedIn drafts and a human-pace queue with random pauses, hourly and daily caps, working hours, and a stop on the first warning. A person presses send; hireburn never does. It also includes a reply and follow-up tracker.

funnel reports losses per stage, show-up and completion rates, days to first work, and the AI label against who is still working on day N. It also includes a candidate disclosure checklist measured against NYC Local Law 144, the Illinois AI Video Interview Act, Illinois HB 3773 and the EU AI Act (orientation, not legal advice).

Install and run

pip install hireburn

Cite

Arsentev, E. hireburn: an open-source AI HR hunter — record-once avatar interviews, human-pace outreach and hiring-funnel loss accounting. DOI 10.5281/zenodo.23979629

Method report: 10.5281/zenodo.23927437

More: https://arsentev.ai/employees · https://arsentev.ai/research