AI signal monitoring for executive search: 7,394 headlines down to 164 real events
An executive-search firm needed to know first when a target company changes its leaders, owner or plants. We built an n8n pipeline with Claude that reads the news and the job boards for 1,145 companies and keeps only what matters, each event with a source and a quote.
All case studies- Client
- a RevOps agency in the DACH region, building for an executive-search firm
- Service
- AI automation, n8n
- Markets
- Germany, Austria, Switzerland, Czech Republic, Slovakia
- Built on
- n8n, Claude Haiku, Google News, LinkedIn jobs via Apify, Baserow
- Status
- full pipeline tested; the 14-day schedule starts next
The situation
In executive search, timing wins the mandate. A new CEO, a change of owner, a plant opening or closing, a restructuring: each one is a reason to call.
But these signals are scattered. They sit in news sites, company press pages and job boards, across five countries and over a thousand companies.
Reading all of that by hand does not scale. Most of it is noise, and the few real events are easy to miss.
What we built
A pipeline on n8n with Claude Haiku at five steps.
- 01Headlines
From Google News it keeps only the headlines about the company itself that look like an event.
- 02Company news pages
It reads the news page on each company's own site and lists every item with title, date and link.
- 03Classification
It reads each article and assigns one of 8 event types: leadership change at the first or second level, change of owner, a plant opened, expanded or closed, restructuring. It adds the date, the people and a verbatim quote.
- 04Duplicates
The same event from several outlets becomes one row.
- 05Jobs
It reads each full LinkedIn job ad and applies the client's own rules, such as management level by country and a salary floor for some roles.
Everything lands in two tables in the client's Baserow: events and jobs. The consultants work from there.
Built to be checked
- Every event carries its source and a verbatim quote, so a consultant can verify it in seconds.
- When the model is not sure, it does not guess. It marks the row for a person to check.
- For jobs, code compares the model's quote with the text of the ad before anything reaches the table.
Results from the test runs
Test 1: 203 companies, a 90-day window.
Test 2: 25 companies.
Speed: 100 companies in about 55 minutes. The full list of 1,145 runs in about four nights. Instead of thousands of headlines, consultants get a short table of real events, each with the proof attached.
What comes next
The full list on a 14-day schedule.
Tracking companies, people or markets by hand?
We start with a fixed-price pilot from $1,000 on a slice of your list, so you see the signal before the full build.
Book a call about your processTracking companies, people or markets by hand?
Tell us in two lines. We'll reply with how an agent would take it over and what the first step would cost.
- A fixed-price first step, from $1,000
- Half at the start, half after we check it on your data
- NDA on request
- [email protected]Vitaliy, CEO
- Vesivärava tn 50-201, 10152 Tallinn, EstoniaHead and development office