The Bank for International Settlements (BIS) is the oldest international financial institution, established in 1930.
Its mission is to support central banks’ pursuit of monetary and financial stability through international cooperation, and to act as a bank for central banks.
BIS maintains a search engine called BIS Data Portal that makes available statistics, compiled in cooperation with central banks and other national authorities, designed to inform analysis of financial stability, international monetary spillovers, and global liquidity.
Some users of the BIS Data Portal have reported that certain specific queries return zero results. BIS therefore intends to address this issue by improving the search engine’s understanding of queries, making them more specific and targeted. Within this scope, they also recognise in the current proliferation of Large Language Models an opportunity to expand the current data portal capabilities.
In response to the BIS requirements, Sease adopted a multi-phase approach to address the issue, beginning with an Exploration & Discovery to assess the current state of their Solr search engine and identify potential improvements, followed by an in-depth study of the requirements and the subsequent design of a Proof of Concept solution that included both development and evaluation.
Over the years, we have refined how to face these scenarios, and the formula we have designed is a set of online and offline activities with the scope of understanding the client search system and giving immediate value through a systematic audit.
The main focus of the Exploration & Discovery phase is to give immediate value to the client, highlighting what they are doing well and what can be done better in terms of processes, architecture, configuration, and usage.
Once team dynamics, communication protocols, and software engineering practices are clarified, it’s time to get technical, so we reviewed the system architecture, Apache Solr schema and solrconfig configurations.
Analysing these aspects is strictly tied to the client’s requirements and how they expect their users to interact with the search engine.
In the specific case of BIS, we focused particularly on analyzing the structure of their queries, explaining how they currently work and how it was possible to resolve the problem they were facing—namely, their inability to obtain the desired results.
We studied the data and the queries, identifying and recommending any best practices in data modelling and building the best Solr query for each of their task. For this specific Exploration & Discovery, we also reviewed BIS’s repository content. This included additional configuration files, scripts, and Python code that interacts with Solr.
The outcome of this first package was a report structured with the following paragraphs:
Each chapter has a short summary with the main takeaways and a list of paragraphs; each paragraph has an additional summary with the key takeaways and a long explanation.
It’s important to make the report readable. Along with the Exploration & Discovery, we carry out various calls to communicate the most important takeaways, but it’s vital to put them down in a clear and usable presentation that is nice for both the business and the technical reader.
The purpose of this Proof of Concept was to demonstrate the feasibility of enhancing the user experience by supporting natural language queries to explore BIS statistical data.
Sease designed and developed the search API, the component that handled search requests from the user interface, interacted with the LLM, and constructed the Solr queries to get results. We implemented it using Python and Flask to develop a RESTful API.
Specifically, we implemented:
The outcome of the POC was both the code and a report highlighting the promising aspects and the weaknesses of the proposed approach and implementation. We also described all the necessary steps to go to production.
We never run a Proof of Concept with the sole scope of convincing a client to go in that direction or chasing the mainstream: we genuinely design and implement the best solution in budget and evaluate thepromising aspects and weaknesses, offering a detailed plan of what it takes to bring it to production, if interesting enough.
In this project, Sease developed a service that suggests filter selections in the BIS data portal UI, starting from the user’s natural language query, exploiting Large Language Models.
The design and implementation consisted of:
The final service flow consisted of:
Also, in this case, a detailed report has been given as a deliverable. Containing for each point an in-depth explanation of what the requirement was, what was done, and why those decisions were made. This is very important also for the future, so that, even after some time has passed, the company can justify the decisions made, verify whether they still make sense and are valid, and, if necessary, consider new optimizations and additions.
Throughout the project lifecycle, Sease maintained a structured workflow, characterized by clear communication channels (web-call and e-mail) and iterative refinement. Regular updates and consultations ensured seamless collaboration between Sease and BIS, fostering alignment with project objectives and milestones.
Through this collaboration, Sease successfully transformed a critical user pain point—zero-result queries—into an opportunity for cutting-edge innovation. By combining a meticulous architectural audit with state-of-the-art AI integration, we didn’t just patch a search engine; we completely elevated how users interact with complex financial data.
The value delivered to the client spans three critical pillars:
Ultimately, Sease equipped the client with a modernized, intelligent search ecosystem that eliminates user friction, unlocks the power of natural language discovery, and secures their position as a forward-thinking leader in data dissemination.
Every successful AI-powered search project starts with understanding the search system behind it. From search assessments and solution design to Proofs of Concept and production deployments, Sease helps organizations build intelligent search experiences that deliver measurable value.
