Granica

Granica builds an AI efficiency platform aimed at exabyte-scale data: turning large volumes of stored data into material usable for real-time reasoning, while reducing cloud data costs and making enterprise datasets AI-ready. The company's work is grounded in research in information theory, probabilistic modelling and distributed systems, and it applies techniques from data compression and privacy preservation to what it describes as self-optimizing data infrastructure.

Three services sit beneath the lakehouse, optimising data itself without disrupting existing pipelines. Granica Crunch is a compression service that the company says can cut cloud data costs by up to 80%. Granica Screen provides privacy preservation, and Granica Chronicle provides data visibility and observability. Query performance is reported to improve substantially: petabytes of data can be queried as though they were terabytes. Deployments in production run to hundreds of petabytes across enterprises in finance, technology and industrial manufacturing.

The technical domains in play include information theory, probabilistic modelling, distributed systems, data compression, privacy preservation, data visibility and observability, lakehouse optimisation and AI data efficiency. The company positions its approach as a shift from traditional ETL towards a paradigm it calls E∑L, in which information efficiency drives AI adoption, and argues that the next advance in AI will come from more efficient data systems rather than larger models alone.