standards for data management, which increases trust in research findings. This is especially
important in engagement with authorities, and for public accountability towards the CSO
sector, media and academia.
Another aspect of great importance in this project has been the usability for non-technical
users. The interface is intuitive and easy to use, while it is open source, which gives additional
opportunities for tech-savvy users.26 This means that the model is quite versatile, but with
emphasis on useability for CSOs with less technical capacity.
When combined with contextual and other crucial knowledge, this AI-powered detection tool
will be able to significantly enhance efforts to document online gendered hate speech.
You can access the Nuha platform and
explore how to use it here: Nuha
4. USABILITY AND TECHNICAL ACCESSIBILITY
A central component of the project was to ensure the AI-powered research tool would be
useful and accessible for civil society organisations of different capacity, and particularly those
working in the field of gender justice and women’s rights. Many CSOs across the MENA region
operate with limited technical resources and budgets, and small teams. Egypt is unfortunately
no exception. For this reason, it was essential for the tool to be intuitive and easy to integrate
into the current work of organisations, without requiring specialised technical expertise.
Meanwhile it needed all the features expected of university and other researchers. The
usability therefore focused not only on the tool’s technical performance but also on how easily
it could be integrated into the work of practitioners and researchers in and on Egypt.
The interface was designed to be clear and straightforward, allowing users to upload large
datasets, run automated analyses, and view results through different visualisations. After the
introductory workshops, most participating organisations were able to navigate the system with
minimal instruction, demonstrating that AI-assisted monitoring can be made accessible even to
teams with limited backgrounds in data science or machine learning. Having trained users in
the country functions as an additional support, also beyond the time frames for the project.
TRAINING FOR APPLYING THE TOOL
The introductory workshops played a crucial role in enabling organisations to make full use of
the tool. They introduced the basic concepts of AI-based detection, explained how the model
processes text (and even emojis), and outlined the limitations and appropriate uses of
automated analysis. Participants were encouraged to approach the tool not as a standalone
solution but as an extension of and complement to their existing expertise. These sessions
helped demystify the AI-tool, and gave users confidence in applying it to their own activities.
The workshop format also enabled CSO representatives and researchers to share their
26 Project reference group input, September 2025
9