top of page

AI Case Study

SoftBank accelerates due diligence of ARM acquisition by reducing patent portfolio searching from weeks to day using machine learning

Softbank and ARM leveraged Aistemos' Cipher tool, which is powered by machine learning, to do the due diligence on both companies' patent portfolios. The process was completed in just a couple of days, instead of the usually lengthy period of multiple weeks, as the software is capable of reading through documents at very high speed, providing efficiency.


Financial Services

Investment Banking And Investment Services

Project Overview

"When it came to doing the due diligence on ARM’s patents, the usually long and laborious process that can last for weeks on end sped through in just a couple of days. The reason was software powered by artificial intelligence that could read documents at light speed, compared to humans. The searches, which would have taken weeks and cost thousands of dollars with human lawyers, took just a few seconds, says Nigel Swycher, whose startup Aistemos built the patent-trawling tool."
According to Swycher, "the human equivalent of a single search on Cipher costs $10,000 and takes two weeks. 'Our average user does 250 searches a year', he says, suggesting the service becomes cost effective after about half a dozen uses."

Reported Results

Due diligence process time reduced from weeks to days.


"Cipher applies this contextual framework, coupled with the latest advances in machine learning, to provide the most innovative approach to patent analytics."


Information Technology

Business Intelligence


"Softbank bought the powerful chip designer ARM for $32 billion in July 2016."




bottom of page