Find the words that matter.

Thresher helps expert analysts find what they are looking for in unstructured text even when authors are being creative, trying to shape the dialogue, or outright trying to hide.


We are an award-winning start up that develops tools for expert analysts to extract meaning from unstructured data in any language without requiring search logs or ontologies. We believe that combining what computers do best with what experts do best creates powerful value. We have paired the tool with analysts to identify codewords that Chinese citizens used to evade censors, find writings about suicide bombings, boost the language processing capabilities of beginner Arabic speakers, sort Arabic texts by dialect, and list slang used to talk about marijuana online.

As one of two data scientists on a small team in a fast-paced environment, I do a lot: everything from algorithm development, core code implementation in Python, data science and engineering in Python and R, data scraping and wrangling, and substantive research design and execution for client-specific questions. The speed of problem identification and solution implementation, the constant engagement with new and challenging questions, and—most of all—the camaraderie of our team, set Thresher apart.