The starting point
A trading firm looking for quantitative researchers and traders with a strong maths or physics background. International competitions such as IMC, EGOI, IYPT and IPT publish participant lists from previous years.
Going through those lists by hand, however, is extremely time-consuming and inefficient.
The problem
- Inconsistent data formats. Results are scattered across different websites, often trapped in unstructured PDF documents.
- Missing context. A name and a country are not enough. A recruiter needs to know where the candidate studied and where they work now.
- No way to scale. Looking up thousands of names on LinkedIn by hand is months of work.
The solution: an AI sourcing automation
We built an automation that combines LLM-based data extraction with enrichment from external databases.
How it works:
- Extraction with Claude. We deployed an automation that goes through the competition websites and uses Claude to extract data from a range of formats, PDFs included, into one central database.
- Data enrichment. Every record was enriched through Coresignal, based on name, country and university, giving us current LinkedIn profiles and work history.
- Automatic ranking. The system ranked participants on the basis of their work history and the medal they won.
What changed
- From manual work to a finished database. The result is a master table of more than 18,000 participants.
- Prioritisation. Candidates are not only ranked but also enriched with contact details and a score.
- Immediate action. A recruiter can send hyper-personalised LinkedIn messages to the top 50 people straight away, instead of spending weeks searching.
What transfers to your company
- Data trapped in PDFs is not an obstacle. Today’s LLMs, Claude among them, reliably extract structured data even from visually complicated documents.
- Connect AI to APIs. The real value does not come from AI alone, but from wiring it to existing databases and systems such as Coresignal, LinkedIn or your CRM.
- Automate the preparation, not the final decision. AI produces the perfect list, but writing to the top candidates is still a person’s job.
How to bring this into your company
If you want to know what automation like this would look like for you, let us start with a 15-minute call. We go through whether it makes sense in your company and which of our workshops is the best starting point.