the problem
A software company selling into K-12 schools needed the right people to talk to: district leaders and the teachers who run animation and digital media programs. That data is scattered across thousands of district websites, and buying a generic list wasn't getting them there.
what i built
A small, configurable pipeline, proven out on a sample before committing to a full run:
- start from the source of truth: every district from federal education data (NCES), with no pre-filtering that might hide good fits.
- two tracks: district-level decision makers and school-level program instructors.
- enrich, with a fallback: contacts pulled from a sales data API, with a scraper plus an LLM to read district websites when the API comes up empty.
- clean and score: deduplication, email validation, then scoring and tiering so sales starts with the best leads.
the result
The proof of concept validated the workflow and the data sources end to end on a small sample, giving the client a clear, low-risk decision on whether to fund a full run. It wasn't taken further, but it answered the question it was built to answer, quickly and cheaply.