The European Council plays a central role in setting the EU’s political direction and shaping its policy agenda. However, like many large organisations, the Council faces challenges in managing its vast administrative workload, particularly in processing and handling documents. With key responsibilities that include foreign policy, crisis management, and security, it’s vital that the Council operates as efficiently as possible. Faktion partnered up with the Council at a crucial time in their digital transformation journey, with both parties recognising the potential of AI to streamline operations.
Challenges in Managing High Volumes of Documents
Despite the Council's pivotal role, many of its day-to-day operations were bogged down by manual processes that slowed down workflows and consumed valuable resources. The Council identified four specific areas where manual effort was particularly high:
- 1. Text Quality Analysis: Assessing the quality of documents produced in multiple languages was time-consuming and inconsistent. There was no standardised method for evaluating grammar, vocabulary, or overall clarity in English texts.
- 2. Text Classification: Sorting and categorising large volumes of documents — ranging from policy papers to meeting minutes — was done manually. This led to delays, risks of misclassification, and inefficiencies.
- 3. Text Summarisation: Summarising lengthy documents, such as reports and policy briefs, was a labor-intensive task. Human error, inconsistent summarisation styles, and inefficiencies were prevalent.
- 4. Meeting Minutes Generation: Writing meeting minutes was entirely manual, resulting in varying quality and style. The process was time-consuming and often delayed decision-making due to the backlog in producing formalised records.
These tasks, although essential, were inefficient and costly when performed manually, pushing the Council to explore AI as a means to automate these processes. Additionally, data security and governance were crucial throughout this exploration due to the geopolitical importance and sensitivity of the Council’s work. The Council required guarantees that AI vendors would respect data sovereignty, ensuring that sensitive political data was not exposed to foreign entities. This data security requirement influenced the choice of potential AI solutions and vendors.
Addressing the Need for AI-Powered Solutions
The European Council needed to automate these labor-intensive tasks without sacrificing the quality or accuracy of their outputs. Their objectives were clear:
-
Improve efficiency: Reduce the time spent on repetitive manual tasks, allowing staff to focus on more strategic work.
-
Ensure consistency: Automate processes to standardise outputs and reduce the variation in quality caused by human handling.
-
Integrate seamlessly: Implement solutions that could fit into the Council's existing workflow and infrastructure without causing disruption.
-
Maintain data security: Ensure that all AI solutions adhered to strict data governance and sovereignty requirements, protecting sensitive political information from foreign access.
With these goals in mind, the Council partnered up with Faktion to assess the feasibility of implementing AI-powered solutions across the four identified areas, while placing heavy emphasis on data security, governance, and sovereignty throughout the project. An extensive benchmarking study was conducted, comparing custom-built solutions to off-the-shelf technologies, with special attention to where the technology was developed and hosted, favouring European-based alternatives when possible.
How We Automated Key Tasks at the European Council
Faktion conducted an extensive feasibility study, complemented by the development of four separate proof-of-concept (PoC). These PoCs were embedded into an intuitive demo tool, designed to showcase the potential of AI in automating the Council’s manual processes.
- 1. Text Quality Analysis: We developed a tool that evaluates the quality of English text produced by the Council. This tool scores documents based on various metrics, including grammar, vocabulary richness, style coherence, and overall readability. More importantly, it provides actionable suggestions for improvement, ensuring that documents meet the Council’s standards while reducing the burden on employees.
- 2. Text Classification: Leveraging the Metamaze Intelligent Document Processing (IDP) platform, we tackled the text classification challenge. Metamaze uses AI algorithms to classify documents and extract relevant information. In the Council’s case, the platform predicted document types, detected qualifiers (e.g., whether a document was an initial version, revision, or correction), and extracted acronyms by analysing content. The platform also managed references to other publications, which helped the Council efficiently link documents across its vast archives, leading to a significant reduction in the time spent on sorting documents.
- 3. Text Summarisation: Faktion’s summarisation tool was designed to automatically condense lengthy documents, such as policy papers and legislative briefs. The tool uses advanced natural language processing techniques to extract the key points from each document, creating concise summaries that retain the original meaning and intent. This not only reduces the time spent on summarising but also ensures consistency across all summaries.
- 4. Meeting Minutes Generation: One of the most time-consuming tasks within the Council was the manual generation of meeting minutes. Our PoC tool automated this process by converting bullet-point notes from meetings into fluent, well-structured summaries. This significantly reduced the time required to produce minutes and ensured that the final documents were consistent, accurate, and ready for immediate distribution.
Each of these PoCs was embedded within a demo application, allowing the Council to see firsthand how these tools could be applied within their daily workflows. The demo was designed not only to highlight the technical feasibility of AI automation but also to provide a clear vision of how these solutions could be scaled and integrated across the organisation.
Our Approach
Our approach began with an in-depth market study for each of the four focus areas. We conducted a thorough evaluation of existing AI solutions, considering their strengths, weaknesses, and potential for deployment within the Council. This included reviewing state-of-the-art technologies and comparing their applicability to the Council’s specific needs.
For each area, we made formal recommendations on the best-fit technologies and proceeded to develop a PoC. These PoCs were designed with scalability in mind, ensuring that they could be expanded upon and fully integrated into the Council's existing infrastructure. We also focused heavily on user experience, ensuring that the tools were intuitive and easy to use by the Council’s staff.
Our development process was highly collaborative. We worked closely with the Council’s Digital Transformation Team to ensure that the tools met their exact specifications and fit seamlessly into their workflows. Regular feedback sessions allowed us to refine and optimise the solutions throughout the development process.
Conclusion
By automating critical manual tasks, Faktion helped the European Council unlock new levels of efficiency and consistency in their operations. The text quality analysis tool standardised document quality across the organisation, the text classification through the Metamaze platform significantly improved document sorting tasks, the summarisation tool reduced the time spent on distilling lengthy reports, and the meeting minutes generation tool ensured that important decisions were recorded accurately and distributed promptly.
This project demonstrated the power of AI in automating high-volume, manual tasks within large organisations. The European Council’s investment in these AI-powered solutions not only streamlined their operations but also set the stage for further digital transformation initiatives in the future. The success of these PoCs, along with the prioritisation of data security, data governance, and data sovereignty, ensured that the solutions met the stringent requirements of a politically significant institution like the European Council.
As organisations increasingly look to AI to solve operational challenges, this project stands as a clear example of how AI can be effectively deployed to drive meaningful improvements. The success of these PoCs has laid the groundwork for potential full-scale implementation, which could further enhance the Council's ability to manage its critical functions efficiently and accurately.
