Leveraging Generative AI for Impact Assessments in Public Sector Organisations

Public sector and philanthropic organisations work towards impact goals that are difficult to measure and evaluate. These are often non-monetary goals that have intangible outcomes that take a long time to materialise. This includes supporting research and innovation, complex social and economic outcomes, as well as environmental and social challenges. In this context, it can be challenging to attribute outcomes to specific interventions, partly because results also depend on external factors.  

Assessing these activities is often complex and requires extensive data collection, which can be both time-consuming and disruptive. At the same time, public sector and philantropic organisations generate large volumes of qualitative and unstructured data to assess the impact of initiatives (such as progress reports, KPI reporting and final reports). Historically, the size and complexity of this data have made it challenging to extract valuable insights about the impact of policies and funding.  

Generative AI can play a transformative role in this context. With its ability to process and analyse large bodies of text, it offers powerful solutions for managing and navigating vast amounts of information. Input data can be data that the organisation has already collected through their funding processes. Analysing this information with generative AI has the potential to significantly enhance the capabilities of impact assessment frameworks.  

Generative AI enables comprehensive analysis of large amounts of text, identifying key elements with respect to outcomes and impacts. It can distill vast amounts of information into concise summaries, highlighting the most critical outcomes and impacts. By considering the entire document base of an organisation, AI can generate granular insights that are both summative and formative. 

This capability is not limited to a single tool or use case. Instead, it represents a broader shift in how organisations can approach impact assessment, with a wide range of applications across sectors and contexts. 

To meet this opportunitiy, ADC has developed an AI assistant for impact management. This assistant interacts with unstructured and semi-structured project data to generate insights in an accessible way. The modular, agentic RAG (Retrieval-Augmented Generation) approach enables answers to both project-specific and high-level strategic questions.  

As such, the tool supports a wide range of use cases – from evaluating the outcomes of individual projects to monitoring trends across a large number of initiatives. A manager overseeing the distribution of public funds can use it to get information on the results of a particular project, or to explore questions spanning multiple programmes and thousands of projects. The tool can answer specific questions, generate summaries, and track progress over time. This makes it valuable for both summative evaluations and ongoing, formative monitoring. 

The tool enables important insights that would be extremely time-consuming and difficult for an individual to retrieve. The AI assistant therefore plays a key role in large-scale evaluations of public sector and philanthropic initiatives. It can help inform better resource allocation, ensure that initiatives are aligned with strategic goals, and track whether they are likely to achieve the desired impact. 

Ultimately, while public sector organisations often pursue goals that are difficult to measure and evaluate due to their intangible and multifaceted nature, the vast amounts of data they generate offer significant potential for evaluating and maximising impact. Generative AI can play a transformative role in unlocking the potential of this data, providing powerful tools for dynamic reporting, advanced analytics, and predictive modeling. By leveraging AI capabilities, public sector organisations can enhance their impact assessment frameworks, enabling more accurate, timely, and actionable insights to support better decision-making.  

In a time in which many public sector organisations are expected to deliver increasing value to their citizens with fewer resources, it is crucial to understand whether they are on track to achieve their long-term goals. Generative AI offers a data-driven way to assess whether these goals are met efficiently and effectively, presenting an exciting opportunity in the complex landscape of public sector impact assessment. 

By leveraging our deep expertise in data science and AI, ADC partners with public sector organisations to build bespoke impact frameworks that are augmented with AI capabilities. Curious about how your organisation could make use of generative AI to maximise its impact? Reach out to Boudicca Georgii Hellberg (Senior Consultant).