third eye 1.PNG

Powered by an evolving AI algorithm, Third Eye enables doctors to intervene early before it is too late. The AI utilizes data routinely collected in intensive care to predict the risk of organ failure, mortality, and length of stay, in view of guiding doctors in evidence-based decision-making.

The vision

Automatic Machine Learning (AutoML)

Instead of hand designing models which are suboptimal, 3rd Eye designs AI algorithms that automatically design optimal models for them. AI designing AI.


Learning a shared representation across multiple tasks. Just like human learning, 3rd Eye AI models learns from different tasks using this to improve its performance in each task.

Meta learning

Instead of learning from labels conventionally, 3rd Eye's model learns how to learn. Allowing the model to learn dynamically on unstructured data, in an ever changing environment and delivering a bias free model.


The technology

Combining multiple facets of machine learning, ranging from unsupervised representation learning to meta learning, 3rd Eye has developed a truly multi-model system capable of running inferences on test, images and signals in unison. The model enables to make predictions that are put into situational context, just like a human would.

Third Eye Intelligence makes it easy for health systems and hospitals to roll out the system by providing data analysis in the cloud, workflow integrated interfaces and constantly improving models and features. Always developed for our end users needs.