Paste in a news article to see how Backfield turns it into structured data. This demo will only use the first 500 words.
Articles are full of useful information: people, places, quotes, organizations, events, facts and relationships. Different kinds of journalism contain different kinds of structure — legislation has bill numbers, restaurant reviews have dishes, obituaries have life histories. Backfield makes that information usable outside the article itself.
Backfield uses generative AI alongside more traditional tools to identify and structure the information inside a story. It can extract common things like people and places, assign metadata like story format and user need, or look for information specific to your newsroom and coverage.
Extraction is only the beginning. Backfield can resolve an offhand mention of a coffee shop into an address, put it on a map and determine its neighborhood; connect an elected official to a district and that district to a geographic area; or link different mentions of the same person or organization across stories.
Backfield keeps the evidence behind what it produces. Extracted facts can point back to the passages that support them, quotes stay connected to their speakers, and entities retain the context in which they appeared. The result can be reviewed, corrected and traced back to the reporting.
Once it is turned into consistent, queryable data, journalism becomes raw material for building new products and services: neighborhood feeds, coverage maps, source audits, personalized products, better analytics, new advertising products, agentic services and applications we haven't invented yet.