Building an Inventory Management Pipeline in Foundry — Looking for Advice!

I’m working on a Foundry project focused on inventory management and procurement optimization for a fictional company. The goal is to build a data-driven system that tracks stock levels, forecasts demand, recommends the best vendors, and alerts the procurement team when materials hit reorder points.

Here’s the data I’m working with:

  • Consumption history (180 days)

  • Inventory levels (90 days)

  • Materials master (costs, categories, reorder points)

  • Plants (2 sites)

  • Purchase order history (12 months)

  • Vendor communications (quality issues, delays, negotiations)

  • Vendor-material pricing (cost, MOQ, lead time)

  • Vendor master data

Main components I’m building:

  • Pipelines – Transform and join data for inventory health & reorder insights

  • Model Integration – Simple recommendation model for vendor selection

  • Unstructured Data Processing – Extract insights from vendor communications

  • Ontology – Link Materials, Vendors, Plants, and POs

  • Application – Workflow for alerts → vendor comparison → order approval

I’d love your input on a few things:

  • Best way to join consumption and inventory time-series for reorder prediction

  • How to process vendor communication text effectively in Foundry

  • Tips for ontology design to connect vendors and materials cleanly

  • Ideas for structuring the alert-review-action workflow in Workshop

Any examples or best practices would be super helpful

Thanks,

1 Like