Key Technology — US-HQ AI-driven optical sorting for NA food processing (Walla Walla WA; 15+ named customers)

NA-US (HQ Walla Walla WA; customers across CA, WA, MT, WI, plus Quebec Canada)

Content

Key Technology, a Duravant subsidiary headquartered in Walla Walla, Washington, builds AI-driven optical sorting and digital inspection equipment for the food-processing industry. Their VERYX belt- and chute-fed sorters, Optyx, Cayman, BioPrint hyperspectral, VitiSort, COMPASS, Taurys brands all sit under one platform.

VERYX 2.0 — the AI-load-bearing product. Key Technology explicitly markets VERYX 2.0 as “powered by artificial intelligence (AI)” — using AI in:

• FM Alert (image classification of foreign material)
• Sort-to-Grade (clump separation for size grading)
• Discovery IIoT analytics platform

Named NA customer deployments (15+, all from Key Technology’s customer-stories index):

What the AI does:

• In-air inspection with Pixel Fusion across multiple sensor types
• AI-assisted foreign-material classification
• AI-driven clump separation for size grading
• Hyperspectral defect detection (BioPrint)

Why this is structurally distinct from TOMRA. Both are optical-sorting vendors with NA presence; Key Technology is US-headquartered and US-engineered (Duravant), while TOMRA is Norway-parented with Sacramento CA US HQ. Talk-stage worth distinguishing — the data-sovereignty considerations are different.

What this unit is doing in the taxonomy

Anchors the processing × computer vision × foreign-material-and-defect-detection cell with a US-HQ comparator to TOMRA’s EU-parent pattern. Sister units:

• TOMRA (tomra-na-food-sorting.md) — EU-parent / Sacramento NA HQ counterpart
• Cargill CarVe (cargill-carve-meat-processing.md) — beef CV, different protein
• Apeel × RipeTrack (apeel-ripetrack.md) — post-harvest, not processing line

Why it matters for talks

• 15+ named NA customers across multiple protein / produce / nut sectors is the strongest NA-deployer footprint in the optical-sorting AI category. Worth surfacing because it makes the point that NA food processing is not under-AI’d; it is well-AI’d in this product class, just rarely named publicly.
• The cross-sector breadth (berries, nuts, leafy greens, pasta, wine) is unusual. Most NA food-processing AI vendors cluster in a single sub-sector; Key Technology’s breadth indicates a vendor with category-spanning platform architecture.
• Chateau Ste. Michelle + Jackson Family Wines + Chappellet — three NA wineries using AI sorters. Wine grape sorting is small in volume but high in value; AI sorters in NA wine are a deployable-yet-underexamined cell.
• The Que/Quebec Wild Blueberries deployment makes this unit cross-border NA-relevant, not just US.
• Hardware-IO + AI building block pattern. Key Technology’s products run on-premise at the line; the AI classification happens at machine speed. Cloud-mediated AI is not the architecture here.

Critical context

• All deployment evidence is from Key Technology’s own customer-stories index — strongest when confirmed by trade press but vendor-reported for the deployment count itself.
• Customer list is not exhaustive; Key Technology has more customers than the 15+ named.
• “AI-powered” framing comes from Key Technology’s own product marketing. The ML model architecture is not publicly documented.
• Foreign-material detection on-tree-nut and berry lines is a high-stakes application (allergens, contaminants). Independent food-safety audit of Key Technology equipment is not surfaced in our sources.
• Duravant ownership: Key Technology is a subsidiary; corporate strategy decisions (potential M&A) could affect US-HQ continuity.