Case 2: AI-SHELFVISION

About Customer

The customer is a multi-store retail chain managing diverse product categories across urban and semi-urban markets. Their merchandising teams are responsible for planogram compliance, stock visibility, and store-level execution quality.

With high SKU movement and frequent promotions, the retailer needed a scalable way to monitor shelf performance consistently across locations without increasing manual audit workload.

Problem

The retailer lacked a dependable way to monitor shelf execution at scale. Day-to-day merchandising decisions depended on manual aisle walks and checklist audits, which consumed store time but still failed to provide timely, consistent visibility into shelf health.

• Out-of-stock conditions and facings issues were often identified too late.
• Planogram non-compliance persisted due to inconsistent field validation.
• Regional and category teams lacked a near-real-time view of shelf reality.
• Fragmented audit records delayed corrective action and replenishment response.

As a result, replenishment and planogram corrections were reactive instead of proactive, creating avoidable lost sales and uneven in-store experience. The business needed continuous shelf intelligence that could convert visual shelf events into clear, operational actions across locations.

Solution

Execution Pipeline

• Step 1 - Data onboarding: We mapped shelf cameras and audit streams to aisle, category, and SKU context for structured store intelligence.
• Step 2 - Compliance extraction: Planogram rules were converted into measurable shelf conditions (facings, position, adjacency) for automated checks.
• Step 3 - CV validation: Models continuously identified OOS events, facings deviations, and display inconsistencies at shelf level.
• Step 4 - Optimization workflow: 3D simulation validated planogram changes before rollout to reduce field execution risk.
• Step 5 - Action loop: Exception dashboards assigned priorities and tracked closure across stores and regional teams.

Conclusion

AI-SHELFVISION improved shelf availability, reduced manual audit effort, and created a scalable merchandising operating model that proved expansion-ready after pilot success.