Skip to content
FelgorFELGOR
Visual inventory counting•Operations with high volume, photo-countable inventory

FELGOR Count

Counts inventory by photographing the stack — anything you count by the unit

The problem

Inventory in large warehouses is done once a year and takes a week. Operations halt, a large team counts unit by unit — and still gets it wrong. The rest of the year the company runs with a gap between official inventory and reality, only finding the discrepancy when a customer wants to place an order and the material doesn't match.

The solution

FELGOR Count replaces manual counting with photos. The operator opens the app, points at the stack (anything: tube, sheet, bag, bale, roll, box), picks the product, takes the picture. The system replies in seconds with the estimated count and an annotated image — every item identified and counted. Operator validates. Done.

Capabilities

What the product does

  • Counts any photo-countable stack: tube, sheet, bar, bag, bale, roll, box, packed piece
  • Works for long, flat and bundled product — auto-learns new formats
  • Annotated image highlights every item for the operator to verify
  • Operator corrects if they disagree — the correction trains the model on the company's context
  • Dashboard of counts by location, operator and product
  • Detects abnormal variation between consecutive counts and alerts automatically

Where it applies

It's not a vertical product

The engine is the same. What changes is the catalog trained for each context. Below are the sectors where we've already seen it make sense — but the use cases don't stop here.

  • Steel, aluminum and metal profile distribution
  • Lumber and construction materials
  • Food industry (bagged goods, packaged commodities)
  • Paper, polymer and chemical logistics
  • Textile and apparel (bales, rolls, packed pieces)
  • Animal feed and agribusiness
  • Glass and ceramics

How it works

Day to day

  1. 1

    Operator opens the app on the warehouse phone

  2. 2

    Photographs the stack, picks the product from the list

  3. 3

    System returns the count in seconds with annotated image

  4. 4

    Operator validates or corrects — the correction stays on file and improves the model

Gains

What changes in the operation

  • Inventory stops being an annual event — can be done as often as needed
  • Operations no longer halt for counting
  • Official inventory matches reality year-round
  • Model gets sharper with every use, specific to the company's products

Who this is for

Any operation with high-volume, photo-countable inventory and chronic drift between official records and reality. The engine isn't tied to one material — the same technology counts steel tubes, feed bags, fabric rolls, paper bales or plastic part packs. Only the trained SKU catalog changes.

Works best when:

  • Product is organized in stacks, pallets, bales or visible units
  • The operation has an operator available to take photos (not a specialist inspector)
  • A defined SKU catalog exists for the operator to pick from

If the product is loose bulk (powder, liquid, unpacked grain), the product doesn't apply — it has to be visually countable.

What FELGOR Count isn't

Not a barcode scanner. Material that arrives without tags is still counted. Not a WMS. The company's existing inventory system stays in place — Count feeds it with reliable counts. Not magic. The operator stays. What changes is what they do with their time: they stop counting, they start validating.

How it fits in

The app installs on operators' phones. The product master comes from the existing ERP. In the first weeks the model learns the warehouse-specific stacking pattern. After that, the operator counts a stack in seconds, not in hours.

Want to learn more about FELGOR Count before booking?

Read articles on this solution

Got a problem similar to FELGOR Count in your operation?

In 30 minutes with the founders, we'll map your context and tell you whether an adaptation makes sense — or whether a different approach fits better. Straight talk.

Book a diagnosis for my case

30 min · free · with the founders