Startup aims to use AI to spot connections in published research

A London-based artificial intelligence startup raised $1 million in seed funding to develop automated “machine reading” programs that aim to digest and distill the expansive mass of documents in biomedical literature and research.

Causaly’s software hopes to process millions of documents, connect the relevant dots, and provide answers to relatively simple questions in the form of cause-and-effect knowledge graphs. The company is already boasting partnerships with Novartis and Harvard University, while its seed round was led by Marathon Venture Capital.

As an example, a researcher looking for the reasons obesity is related to cancer risk may have to hunt for information from several different fields and journals. Instead, Causaly’s semantic search processes and analytics aim to allow for instant gathering of evidence for hypotheses in drug discovery, health economics and pharmacovigilance.

“Getting to the question of how and why things work is at the core of making hypotheses and decisions,” Causaly's co-founder and CEO, Yiannis Kiachopoulos, said in a statement.

The company’s platform is designed to deconstruct and turn free-flowing text into graphs, and hopefully predict any missing links in the mechanisms of diseases. Causaly’s programs have already extracted more than 100 million associations and relationships between terms from published academic literature, the company said.

"Causality used to be a philosophical term. Causaly makes it practical, putting it in the hands of researchers and decisionmakers,” said George Tziralis, partner at Marathon Venture Capital.