The ALFIE project recently hosted its Advisory Board meeting on Monday, the 20th of April, marking an important moment for our development team. Bringing together experts, researchers, and project partners, this session provided an invaluable platform to evaluate our progress on the ETD-Hub and AutoML interfaces.
As we work toward making our tools both powerful and accessible, the insights gathered from our board members have helped refine our roadmap and sharpen our focus on real-world impact.
Shaping the ETD-Hub
A significant portion of our discussion centered on the ETD-Hub, our platform for knowledge sharing and expert collaboration. Key takeaways included:
- Human-in-the-Loop: We clarified that the system currently relies on human expertise rather than AI-generated responses. This ensures the high accuracy and nuance required for complex professional queries.
- Knowledge Management: We are leveraging a semantic knowledge graph to organize Frequently Asked Questions, making it easier for users to find the information they need efficiently.
- Usability & Inclusion: The board emphasized the importance of discoverability and accessibility. We are committed to ensuring the platform is indexed for search engines, is fully compliant with accessibility standards.
Advancing AutoML Capabilities
The demonstration of our AutoML web interface prompted a deep dive into the technical evolution of the tool.
- Refining Models: We received constructive feedback on algorithm selection and data handling.
- Accuracy & Trust: Our participants highlighted the importance of bias report accuracy. We are prioritizing rigorous testing of these reports to ensure that users can trust the automated insights the system provides.
- Data Integration: Discussions around RAG (Retrieval-Augmented Generation) systems confirmed that we are on the right path toward integrating external data points to enhance search results and user understanding.
Real-World Application: The BOSCH Use Case
We were pleased to present a practical application of our tools through our partnership with BOSCH. By utilizing both AutoML and ETD-Hub, ethical officers are better equipped to manage high volumes of compliance alerts. This use case highlights our core mission: using technology not to replace human decision-making, but to act as a powerful filter that empowers experts to make faster, more informed choices.
Looking Ahead
The Advisory Board underscored the importance of transparency, communication, and iterative feedback. We are eager to adopt these suggestions, including a stronger focus on user-driven features.
We want to thank our Advisory Board for their expertise and their commitment to the ALFIE project. Their guidance is instrumental in ensuring that our final tools are not only technically robust but also genuinely useful for the communities they serve.
Stay tuned to the ALFIE blog for more updates as we move toward our next development milestones.