Skip to main content

NATO works on artificial intelligence with the project AI FELIX

AL FELIX assists NATO staff

Published on: 12/08/2021 News

AI FELIX is a modular and agile project based on open source technologies and built by NATO, with the intention of reducing workload for internal staff. In the future, it may be used by member states and embassies for the Alliance, where decisions will be based on vast amounts of information.

 

The Artificial Intelligence Front End Learning Information Execution, AI FELIX in short, is a NATO experiment with (open source) AI for internal use. AI FELIX processes incoming correspondence and documents received in order to extract and analyse key information within these, with the aim of reducing the manual staff effort for monitoring. The second version of the project was presented in 2020.

 

Correct information to correct decision maker

AI FELIX is made for automatising metadata from internal documents and external information received by NATO. It then distributes this information across the organisation. By using this technology, NATO not only saves a lot of man hours, but also ensures that the right information is sent to the right decision makers. If the tool were to be distributed to other NATO organisations, it is estimated that workload would be reduced for up to 40 personnel. AI FELIX has consistently delivered highly accurate results with an average processing time of 27 seconds per document compared to the 5 to 10 minutes required by a person.

 

Deliverables of AI FELIX

The project has a Graphical User Interface where users can link AI FELIX with information and knowledge tools like Enterprise Document Management System, Tasker Tracker Plus and the NATO information Portal. This is not only efficient, but the technology also continually learns more and more about where and to whom the information goes.

NATO highlights three major functionalities and modules for AL FELIX:

  • Command Review Board/Tasking Module: Thanks to this module, the AI can read incoming mail that arrives to the headquarters every day and recommend how to distribute it internally.
  • Portfolio Management Module: Real time and holistic. The AI supports with planning, execution, and delivery of the output more proactively. Among others, it monitors schedules, sends reminders if staff is falling behind, and informs of risks.
  • Knowledge Wingman Module: AI FELIX scans the NATO enterprise for similar work, potential points of contact and generates a community of interest. This networking element is meant to facilitate collaborative stakeholder engagement at the earliest opportunity to support delivery of outputs. It predicts training and education resources to facilitate the staff’s delivery of their primary and associated roles.