Good morning, all.

We're looking for collaborators for the TEA platform and some associated projects (e.g. CVD-Net, Project Bluebird, DTNet+).

If you are interested in any of the following, please let us know, either in thread or via DM:

1. 🛠️ **Techniques Dataset for Trustworthy and Ethical AI:** a structured repository of 125 (technical and non-technical) techniques for trustworthy and ethical AI (e.g. occlusion sensitivity, variable importance in random forests, LIME/SHAP, red teaming, and stakeholder engagement). We've got a web app and dataset that is now ready for user testing, and we'd be keen to publish a paper on the resource once it's ready.
2. ✅ **Assurance and Evaluation of a LLM-enabled pipeline**: while we intend to build community feedback mechanisms into the above platform to help ensure resources remain up-to-date, because of the size of the dataset it was also helpful to have an automated way to fetch relevant resources from sources such as arXiv, CORE, GitHub and Semantic Scholar. An LLM is currently used at a key stage of this pipeline (currently either Claude Sonnet 3.7 or Gemma 3), and there is the potential for doing some empirical verification/validation of the pipeline's capabilities based on the different models. The current pipeline is a series of Python modules, and is amenable to reproducible experiments.
3. 👆**Case Studies and Capability Building**: the CVD-Net and Project Bluebird teams are already building their own assurance cases, using the TEA platform. It would be great if there were other projects across the Turing who would be willing to use their project as the basis for a case study. We're in the process of adding a new community feature to the TEA platform to allow individuals or teams to showcase their work, and are happy to run skills/capabilities workshops for project teams through the [Turing Commons](https://alan-turing-institute.github.io/turing-commons/).
4. 🏢 **Community Building Events and Activities**: over the coming months we will be planning a series of workshops with stakeholders across the public and private sectors, including the Department for Science, Innovation, and Technology (DSIT)—who featured this project in their policy report last year on AI assurance—and the Ada Lovelace Institute. If you are part of a project team working on a topic related to assurance that would would like to be represented in these discussions, please reach out.

:::success
**:link: Links to Resources**
- Trustworthy and Ethical Assurance Platform (an online set of tools and resources for AI assurance)
- GitHub Repository: https://github.com/alan-turing-institute/AssurancePlatform
- Online Platform: https://assuranceplatform.azurewebsites.net/
- TEA Techniques (a community-focused and interactive database of techniques for helping individuals, teams, and organisations build trustworthy and ethical systems)
- GitHub Repository:
- Online Database:
- TEA Resource Pipeline (a helper tool to systematically search repositories for relevant resources to enhance the techniques database)
- GitHub Repository:
:::