Phoenix is Build Up’s process for peace and mediation practitioners to work ethically with social media data to inform programming. It is social media analysis made by and for peacebuilders.

We apply participatory action research to social media analysis, in five steps:

Problem statements that address peacebuilding or mediation objectives

Data pipeline to add the social media sources you need

Customisable classification models to reduce manual organising of data

Dashboard to visualise engagement, classifications and networks

Recommendations & interventions that respond to social media insights

Our vision

  • Create accessible process guides and open-source tools for peacebuilders and mediators around the world to be able to do social media analysis, contributing to decolonise data
  • Create software and AI systems that can provide new insights into polarisation on digital media, including social media
  • Create a community of developers and peacebuilders that can sustain the longevity, and demonstrates the possibility for future collaborations

The tools

Phoenix is currently an open-access public manual and a set of open-source, non-commercial, customisable tools built in partnership with datavaluepeople. You can view the open-access public manual here. You can download the code for the tools by visiting our GitLab (see the link below) and run the tools on your computer. There are four tools available for download:

  • Phoenix-gather helps you to collect data using your own API keys from Facebook, Twitter, Instagram, YouTube
  • Phoenix-tabulate to move the data into a table, anonymise it, and organise it
  • Phoenix-classify to add a classification (label) to each row of data
  • Phoenix-visualise to connect to a dashboard in order view the data in graphs and tables

In 2024, we are working to build Phoenix v.2, which will be a  self-deployable platform that anyone can request a login for, and use with no need to download code, use a terminal or have your own API keys. With some limits to how much data you can process, Phoenix v.2 will be free of charge. Get in touch below to hear more about this development.

Get in touch

We have collaborated with UN agencies, local peacebuilders, academic institutions and researchers to use Phoenix in 20+ countries. Do you have an idea for collaboration or questions about using Phoenix? Get in touch!

Access the code

The Phoenix code is open source & non-commercial. You can set it up if you want using the documentation on out GitLab. Phoenix code is licensed under AGLP-3.0.


The Automatic Classifications for Peace group helps to further the development of open-source text classification systems for peacebuilding and mediation. If you are working on software for text processing (NLP), language research or building open-source tools for social media analysis and are interested in joining others doing same, see how to get involved here.