At Digital Catapult, we engaged Livedata Limited to conduct research on two key areas of Natural Language Processing: abstractive text summarisation and extractive text generalisation. Through extensive research and testing, Livedata Limited identified algorithms that efficiently provide insights into the contents of a passage. They also adapted existing methodologies to automatically remove sensitive information about individuals and companies. As a result, Livedata Limited developed an algorithm that effectively imitates human speech while limiting access to confidential information without distorting the original meaning.

During our collaboration with Livedata Limited on a joint project for the UK Environment Agency, they spearheaded the development of a system aimed at automatically detecting changes in water levels across the UK. Their expertise enabled the creation of a model using Machine Learning, AI, and Big Data analysis to identify baseline water levels from CCTV footage. This innovative approach allowed for early warnings to the population about potential floods up to two weeks in advance, accurately estimating the probability and severity of impending flood events.

During our collaboration with Livedata Limited, they played a crucial role in embedding quality into our production processes and ensuring that our final products meet all key requirements. Livedata integrated process maps into our QA system, significantly enhancing productivity and minimizing operational losses. They also provided comprehensive training for all staff involved in the production cycle, focusing on efficient project management and code writing practices. As a result, the long-term supportability and versatility of our code improved notably, supported by a robust library of standard functions that accelerated our development processes.