RiLP - AI Driven Operational Optimization

Integrated AI platform to create specialized AI models built to make optimal resource scheduling/allocation decisions in manufacturing and supply chain domains.

Manufacturing and supply chain environments are typically highly dynamic (customer orders are received daily or hourly, machines sometimes fail or must stop for maintenance, etc.) and therefore there is no simple way of creating an optimized plan in real time for processes such as manufacturing job scheduling or vehicle dispatching.

RiLP take advantage of a specialized set of machine learning algorithms known by Deep Reinforcement Learning (DRL) that learn the best of sequence of actions to achieve an optimal outcome, precisely what’s required for resource scheduling or allocation, to automate or support the decision process.

The main steps to create and deploy an AI agent are:

  • Define a baseline performance the AI agent should be compared.
  • Configure a RiLP simulator SDK template to train the AI agent.
  • Create one or more clusters of virtual machines in in any of the major computing cloud platforms (Azure, AWS, CGP) to provide high levels of scalability for the training process.
  • Run train trials, adjust the simulator, and tune the DRL algorithms until a good enough performant AI agent is obtained.
  • Deploy one or more agents on a virtual machine or on one of the created clusters created.
  • Once deployed calls to the AI agent can be made through HTTP / REST API.

Key differentiators



Main Use Cases


  • Manufacturing
Automate machine scheduling to maximize throughput when faced with input and demand variability


  • Pharmaceuticals
Automate chemistry, manufacturing, and controls (CMC)
to maximize batch yield and quality 
  • Transport & Logistics

Optimize routing, logistics network planning, and warehouse operations to reduce costs and improve customer satisfaction

Optimize inbound and outbound delivery networks to minimize shipping delays and associated costs


  • Retail
Deliver advanced personalization to adapt promotions, offers, and recommendations for increased customer and sales


  • Energy

Minimize the operating cost of microgrids considering the uncertainty of renewable energy generation load demand, and electricity prices

Optimize predictive maintenance interventions to reduce costs (unplanned failures, unplanned outages, repairs…)


  • Oil & Gas
Optimize tanker routing to reduce costs and ensure on-time delivery