how it works

Apply state-of-the-art reinforcement learning techniques to maximize desired outcomes from existing simulation models.

Import Existing Simulation Model

Create Reward Function

Apply Reinforcement Learning

Boost Compute in the Cloud

Export Optimized Policy for Model

Before

Fixed Value

An optimized value identified to achieve the desired outcomes of throughput and wait times via individual experiments, manual tuning, and runs utilizing local compute resources.

After

Trained Policy

An optimized policy, comprised of different values in time, created to maximize throughput and minimize wait times via parallel experiments, automated tuning, and on-demand cloud compute.

Pathmind

Simulation learning in the cloud.

+
automatic comparisons
+
built-in neural networks
+
guided creation and validation
+
ready-to-export
policy files
+
cloud compute

EARLY ACCESS

Sign up to learn more and join private beta.