Challenge

This challenge responds to the needs described below:

  • How can we explore models for the automatic extraction and structuring of information on the effect of the subsidies on the supported SMEs, from different sources of information we already have available (PDFs, spreadsheets, etc.)?
  • How can we visualise and analyse the information in a way that facilitates drawing conclusions, and to automatically identify errors or deviations?
  • How can we identify, compile and visualise related external data (sectoral, economic, social, among others), that complete and feed into the data?

Context

The Department of Economic Promotion of Bizkaia Provincial Council (BPC) aims to strengthen Bizkaia’s business ecosystem, for which it offers a wide range of grants to promote the competitiveness of SMEs, in areas such as adaptation to a digital economy, attracting and retaining talent, supporting competitiveness at regional level and innovation.

They currently offer grants and services to achieve objectives such as internationalisation, innovation or sustainability. In the process, information is collected from the beneficiary SMEs through evaluation and monitoring reports and on-site visits.

With the aim of helping to measure the impact that the grants and services have on SMEs, we consider it necessary to migrate the data already available to automatic models that offer visualisation, better use of the information and generate historical data to draw conclusions and make decisions. Focusing the work on a specific, long-running subsidy programme.

Opportunities

Three main spaces of opportunity are observed where the application of new ideas and digital solutions could help the Economic Promotion team:

  1. Extract data from current documents (PDF and other formats) and migrate them into more elaborate models.
  2. Explore what other data from open sources can enrich our own data.
  3. Facilitate an understanding of the data and its use for decision-making and impact measurement through their visualisation and automated analysis.

Challenge Description (PDF)