Our client sought to revolutionise its data capabilities by creating a state-of-the-art tool, placing data and technology at the core of its operations. Existing data efforts were often laborious and posed significant operational risks. Additionally, the over-utilisation of analytical talent limited the department's effectiveness, putting it at risk of falling behind.
The project aimed to equip the client with rapid, comprehensive geopolitical information in an easily digestible format. The initial goal was to develop a proof-of-concept tool to demonstrate the value and applicability of generative AI for analysing international relations using open-source data.
QuantSpark was selected for its extensive experience in bespoke data solutions, its commitment to creativity and innovation, and its deep in-house subject matter expertise.
QuantSpark developed a pipeline to source unstructured data and process it through an LLM-engine, extracting and storing structured information in a relational database. Simultaneously, we designed a bespoke web application based on product-led workshops with end users. This application enabled users to create custom visualisations, view raw tables, and access original source texts for each data point.
Using its design-thinking, product-led approach, QuantSpark deployed its interdisciplinary team to deliver a proof-of-concept in just a few weeks. This included experts in Machine Learning, Analytics, Cloud Engineering, Software Development, Product, Graphic Design, UI/UX, Commercial Strategy, and Strategic Insights. We collaborated closely with stakeholders to identify the most immediate opportunities, focusing on enhancing a single aspect of their operations.
The infrastructure delivered by QuantSpark processed over 1 million pieces of content within hours, which would have taken an individual approximately 9.5 years of continuous effort to process manually. Our test scenarios demonstrated that the tool could replicate the same geopolitical insights as existing manual data gathering.
Additionally, the tool reduced operational risks by employing secure and indirect data handling techniques. The dynamic front end allowed non-technical users to ask questions and communicate answers downstream effectively.
Our proof-of-concept tool is the first of its kind for this client, introducing a new data-centric capability and the potential to enhance their current functions with increased efficiency and novel insights.