Creating transparent relevancy in artificial intelligence.
EdgeAi is healthcare's first operational embedded AI with Explainability (xAi). EdgeAi exposes the internal mechanics of machine learning in human understandable terms.
If you were to ask your healthcare staff what makes their AI applications relevant, you would probably receive blank stares. Trust in AI is paramount, but trust is based on knowing and understanding the inputs.
Explainability: EdgeAi offers insight into which features influence predictions and forecasts. Explainability leads to understandability.
By rendering the inputs transparent, you bring trust throughout your organization.
EdgeAi automatically trains models using both historical and recent datasets. Models are regularly compared against previous versions, and high-performing models automatically replace lower-performing models. EdgeAi automatically incorporates novel models into future training and evaluation.
EdgeAi uses your data to create your AI algorithms. Generalized models trained on "mystery data" from other health systems are poor substitutes for your patient populations. Models trained on your data adapt as your population's circumstances, demographics, and other details change.
The Edge Cognitive Platform (ECP) engine automates real-time training-data curation.
EdgeAi in Action
EdgeAi is used in Discharge as a Service®️ to predict anticipated discharges. From actionability to frequency, EdgeAi facts are displayed to create transparency.
EdgeAi forecasts the total number of patients to be discharged by the end of the current day using machine learning algorithms trained on your historical data.
Predictions are unit-level and aggregated to the hospital, region, or enterprise level.
Anticipating and surfacing patient flow bottlenecks
Our Platform and EdgeAi
We’ve built our Edge Cognitive Platform (ECP), the system that powers our operations centers around the capability of EdgeAi. EdgeAi is possible due to how we think about data.
Data as a Cognitive Stream
ECP views data as a stream - from this stream the engine ingests data applicable to specific outcomes.
Real-time processing and real-time presentation
Next-Generation Embedded AI
“Inline computing”: Real-time time stock market trading technology
”Stack” Machine Learning (AI) algorithms enables high performance
Transforms data into “Streams” for outcomes definition
High Performance Cache of Transactional Data for Prediction and action