All Success Stories
Intel VCaaS
Intel

Intel VCaaS

Retail & eCommerce

AI-powered edge checkout system with real-time produce recognition. Reduced customer wait time from 35 to under 10 minutes with automated model updates and operational analytics.

Intel VCaaS is an edge-deployed AI vision checkout system built in partnership with Intel. It identifies fresh produce in real time using computer vision, integrated with live pricing for seamless automatic checkout — eliminating the need for manual cashier identification and dramatically reducing wait times in retail stores.

65%+Checkout time reduced
<100msInference latency on edge
0Cloud dependency for real-time ops

Key capabilities

Real-time produce recognition

Computer vision identifies fruits and vegetables instantly — no barcode, no manual lookup, no cashier intervention required.

Edge deployment

Runs entirely on minimal in-store hardware with <100ms inference latency. No cloud dependency for real-time operations.

Automated model retraining

Scalable ML-Ops pipeline with automated model retraining on edge devices — the system improves continuously without manual intervention.

Operational analytics

OpenTelemetry observability layer gives store managers real-time analytics on checkout performance, throughput and system health.

Edge AI with full ML-Ops and observability

The system runs vision inference directly on Intel edge hardware — no cloud round-trip for real-time operations. A full ML-Ops pipeline handles data collection, model training, validation and deployment to edge devices automatically. OpenTelemetry provides end-to-end observability from inference latency to business metrics. The architecture is designed as a complete retail SaaS — multi-tenant, horizontally scalable, deployable across store networks.

Kenpath's role

Kenpath partnered with Intel as the technology partner, focusing on building the scalable ML-Ops pipeline for automated model retraining, the data annotation infrastructure, operational dashboards for store managers, and the OpenTelemetry telemetry and observability layer. The team delivered production-grade ML-Ops, annotation workflows and real-time analytics — turning Intel's Edge AI and vision models into a deployable retail SaaS.

Built withComputer VisionEdge AIMLOpsOpenTelemetryPythonKafkaCloud SaaS
Read the full success story