The problem
The language was all there, but in PDFs that couldn’t be searched together. Comparing how different contracts handled one issue meant opening them one at a time.
What we built
We took the rep’s library and built a search for it. Every contract was read and split into articles, so the rep could search by topic or exact wording, filter by union, employer, state or article type, see the matching language highlighted in context, and click straight through to the original document.
More than keywords
Keyword search finds exact words and phrases. Semantic search finds language by what it means, so a search for “mandatory overtime” also turns up contracts that say “required to work beyond the scheduled shift,” however each one words it. AI-assisted search goes a step further: ask how a set of contracts handles an issue, or how their wage scales compare, and get a quick side-by-side comparison of the language or the rates, run on our local AI.
Exporting what you find
Any article the rep identified could be exported on its own as a PDF or Word document, ready to share or drop into a proposal. Articles checked off across a search could be exported together as a CSV spreadsheet.
What it meant for the rep
Instead of reading contracts one at a time, the rep could line up every version of a provision, pick out the strongest language, and bring it to the organization as a recommended model, with the source contract one click away.
