craft ai is based on an “always-on” agents architecture enabling continuous learning. The agents automated workflow will keep processing data as it is sent to craft ai.
Agents learn at the individual level: anything from customer and sensor to whole building or process activity. Each agent computes data and context to deliver adaptive decision models.
craft ai API generates white box models to achieve explainable AI. AI should be easy to integrate and easy to trust, not a black box you have no control over.
craft ai implements an hybrid approach: it learns in the cloud and generates models that are embedded in apps & things. High resilience, no latency, no loss of information.
craft ai retrieves and preprocesses additional context data to provide improved, actionable decision models: raw data finally makes sense.