ECP Analytics Pipeline

Problem

As part of the Eindhoven Career Platform committee, I was pulling data manually each week from Google Analytics, LinkedIn, and Instagram, then compiling it into a report. Copy-pasting numbers from three separate dashboards and turning them into coherent insights took hours per week — time that should have been spent on the platform itself.

Approach

Built a Python scraper that collects data from all three platforms on a schedule. The raw numbers are passed to the Claude API, which synthesises them into a structured insight report: identifying week-on-week trends, flagging anomalies, and writing a human-readable summary ready for the committee. No manual intervention at any step.

Outcome

The weekly reporting process is fully automated. The committee now receives a complete insight report without any manual data collection or writing. Time previously spent on copy-pasting is now spent on decisions.