AI Recruiting Sourcing Assistant

Build shortlists for hard-to-fill roles, with each candidate's fit against the brief written out.

Monthly income

$200–$10,000+

Startup cost

$0–$300

Time required

5–15 hrs/week

Time to first $

2–6 weeks

Difficulty

7 / 10

AI leverage

High

AI tools you'd use

ChatGPT
Claude
LinkedIn Recruiter
spreadsheets

Why it works

Sourcing is the slowest part of recruiting and the part agencies overcharge for, and screening reasoning shown explicitly is what a hiring manager can actually argue with - a ranked list without reasons is unusable.

How it makes money

Per deliverable + retainer

How to start

Choose a role family you can assess: backend engineering, finance, clinical.
Take a proper brief - the must-haves, the nice-to-haves, and what the manager will actually reject on.
Build the search strings for LinkedIn Recruiter, including the adjacent titles.
Screen against the brief and write one line per candidate on why they fit.
Use Claude to summarise profiles, not to judge them - screening on AI inference is where bias enters.
Never filter on anything that proxies a protected characteristic, and keep the audit trail.
Deliver a shortlist of ten with the reasoning visible, not a spreadsheet of two hundred.
Include the ones you rejected and why, which is what shows the search was thorough.
Agree data handling in writing before touching candidate information.
Charge for sourcing per role rather than a placement percentage.
Track which shortlists led to hires and tighten the brief accordingly.

AI Recruiting Sourcing Assistant

AI tools you'd use

Why it works

How it makes money

How to start

Tags