Developing, implementing and fully operationalizing models is nontrivial at best. To facilitate the full lifecycle for augmented intelligence, DIHI is at the forefront of implementing enabling technologies by leading the efforts to implement the Duke Health Inference Platform to provide continuous innovation, integration and learning and value generation. Our environment encompasses many aspects including:
- curating usable health data from various data sources daily and in near-real time;
- model development and training;
- deploying computed predictions and / or phenotypes;
- scheduling and execution of deployed models;
- monitoring of the operational stack and invoked tasks, model performance and data quality;
- support for dashboards and workflows driven by model predictions, and
- persisting model predictions.
If you would like to learn more about our approach, please Contact Us.
