HomeCase studiesA matching engine
Something I built · AI and automation
A matching engine
It reads thousands of public listings, keeps only the ones that fit a written profile, scores them and writes a tailored document for each.
How it works, drawn for this page. No real data.
What I did
A profile in plain words
The rules are written down first: what counts as a fit, what is out, and which sectors to favour. A first, simple filter removes the obvious misses. Claude then reads each remaining listing in full and judges it against the profile.
A score anyone can read
Each match gets a score in five parts: fit, skills, sector, location and budget. The parts are shown on every card, so it is clear why something ranked high or low.
Straight into the dashboard
Matches land in my HQ dashboard with a status, a link and the score. Alerts from other platforms that arrive by email are read and added too, so everything is in one list.
A tailored document for each match
For every match, the engine writes a tailored document using only facts from one source file. It cannot invent anything, and I read every document before I use it.
What came of it
Results.
Figures belong to this engagement and this client. Nothing here is an average or a benchmark, and no number appears without the work behind it described above.
What I took from it
The same pattern fits many businesses: write down what good looks like, let AI do the reading, and keep a person for the decision.
Something similar, at your company?
If I am not the right fit, I will say so on the call and tell you who is.