overview
The solution architecture shows how to optimize supply chain networks by solving for production planning, inbound purchasing, outbound selling, and transportation logistics decisions using Linear Programming.
- Use Faker() library to create datasets for Factory, Distributor, and Customer entities
- Leverage Snowpark dataframes to create views for shipments, distributor_rates and so on
- Invoke Snowflake Cortex LLM SQL functions to enrich the dataset with additional fields
- Using the PuLP package (a Linear Programming modeler) and CBC solver for minimizing distance, minimizing freight costs and minimizing total fulfillment costs
- Build a Streamlit application to serve as a web interface for supply chain decision makers to use this solution