DeepSim automates manual processes in your existing design and optimization workflow, by allowing subject matter experts to guide A.I. algorithms (based on Reinforcement Learning). DeepSim integrates with simulators you already use to optimize results without painstaking manual heuristic design. This automation delivers superior performance and faster time to market by exploring enormous parameter spaces faster than humanly possible.
DeepSim is powered by a scalable cloud platform, giving you fast response times even when scaling up your problem. With the ability to test and validate solutions, and to fine-tune the controllers via real-world input (and hardware in the loop options), you can overcome sim2real problems with ease. Patent pending technology based on novel A.I. methods for hardware and runtime selection makes the platform cost effective and reduces the overall development cost.
The DeepSim platform generates controller software, which is optimized for the final deployment platform to ensure all the relevant deployment KPIs such as latency, throughput and energy use are met.
For more information see the full product description here.