Põhisisu juurde

Diffeo Enterprise

Diffeo, Inc.

Diffeo Enterprise: self-driving search engine for all of your disparate data

The Diffeo AI-powered research assistant

The Diffeo platform uses collaborative search and discovery algorithms to augment the intelligence of human users.

Business Audience

CIOs, knowledge management teams and thought leaders love Diffeo, because it is often the first AI solution that their end users really appreciate.

Customer quote: "I have evaluated hundreds of AI tools, and Diffeo is the first that I see actually accelerating my work." -- Managing Director at a major bank on Wall Street.

Any organization with knowledge workers that must search through their enterprise data to make business decisions will benefit from the Diffeo Engine. All vertical industries that use email and business documents, such as Word and PowerPoint, can accelerate their internal teams by deploying a Diffeo Engine on their internal data.

Target Industry

Key industries with an intense need for Diffeo include national defense, financial services, oil & gas, legal, consulting, biotech, pharma, and many other businesses coping with information overload that hinders their business execution.

Value Proposition

Users invite Diffeo into familiar tools like Firefox and Outlook, so that it can highlight knowledge gaps and uncover connections. Diffeo indexes all of the dark data in your shared drives and email archives, and also mines the deep Web to help research analysts understand the whole story.

Diffeo uses natural language processing to recognize mentions of people, organizations and other concepts in PDFs and pages in your web browser and draft documents in Word, Outlook, Quip etc. By inviting Diffeo into these familiar user interfaces, they become user-facing sensors that enable the machine to understand your current work. Diffeo automatically builds a dynamic knowledge graph of related entities by mining many data sources, including your Exchange email and team shared drives, and automatically formulates queries to lower-level search engines.

History of Diffeo Labs

Diffeo Labs has been studying collaboration as a cornerstone of intelligence since 2012. The recommender engine grew out of DARPA Memex and NIST's Text Retrieval Conference (TREC) where the Diffeo team organized the TREC Knowledge Base Acceleration (KBA) and Dynamic Domain (DD) evaluations from 2012 through 2015.

To measure which recommender algorithms suggest novel content to users, Diffeo has refined protocols for both offline evaluations and task completion studies. In 2016, the Diffeo recommender’s results on these metrics caused it to graduate from Diffeo Labs into production use. Today, Diffeo Labs is re-engineering social intelligence to enable machines to collaborate with people.