Heinen's checkout audit
Workflow Procedure for Heinen’s Checkout Task – Scan Audit
1. Access the Dragonfruit AI Platform
Log into the Dragonfruit AI platform.
Select Heinen’s as the organization.
Navigate to the Checkout Insights tab.
2. Selecting and Reviewing Transactions
Once in the Checkout Insights section, review the list of available transactions.
Select a transaction and open the associated camera footage.
Carefully observe the video to check if all items were scanned correctly by the cashier.
3. Identifying and Categorizing Issues
If there are discrepancies between the video and the transaction record, classify them into the appropriate category:
During the Heinen’s Checkout Task, each transaction is categorized into a priority level based on the nature of the discrepancy between the receipt and the video footage. This classification helps in determining the level of concern and whether further investigation is required.
Priority Classification
🟢 Priority 5 (No Issues) – "Matched Transactions"
✔ The item visible in the video matches the description on the receipt. ✔ No discrepancies found. ✔ The transaction is approved and marked as completed.
🟡 Priority 4 (Minor Discrepancies) – "Relabeling Required"
⚠️ There is a slight mismatch between the receipt description and the video, but no loss is suspected. Examples:
Case A: Receipt shows "White Signature Cake", but video shows a slightly different type of cake.
Case B: Receipt shows "Regular Milk", but video shows "2% Fat Milk".
Case C: Receipt shows "Banana", but video shows "Organic Banana".
📌 Action: These cases require technical relabeling, but no major issue is flagged.
🟠 Priority 3 (Pending Review) – "Incident Unreviewed"
❌ The transaction has not been reviewed yet or was left incomplete.
📌 Action: The audit must be completed before the shift ends to avoid leaving unverified transactions.
🔴 Priority 2 (Unclear Details) – "Needs Internal Review"
⚠️ The item is not clearly visible or cannot be verified for other reasons.
📌 Action: These cases require a manual internal review before determining if further action is needed.
🚨 Priority 1 (Obvious Loss) – "Critical Incident"
🔥 The item in the video does not match the receipt at all, indicating a potential loss or fraud. Examples:
Receipt shows "White Signature Cake", but the video clearly shows a wine bottle.
Receipt shows a regular grocery item, but the video shows high-value electronics.
📌 Action:
Immediately add the incident to the investigation report.
Check the "J" Column: Enter the transaction ID only once and list incidents in serial order in the "I" column.
Look for Suspicious Comment Patterns: If the number of comments is significantly higher than the number of items (e.g., 34 comments for a 6-item transaction), it may indicate possible loss or fraud.
Key Notes for Supervisors & Executives:
✔ Ensure transactions are classified into the correct priority level. ✔ Flag all Priority 1 incidents for further investigation and reporting. ✔ Double-check unclear transactions before marking them for internal review (Priority 2). ✔ Complete all reviews before shift handover to avoid Priority 3 (Pending Review) cases.
This priority system ensures structured and accurate auditing while helping to identify potential loss cases quickly. 🚀 Let me know if you need any refinements!
4. Documenting Findings
Record findings directly in Dragonfruit AI’s Checkout Insights section for each reviewed transaction.
If any discrepancies are found, ensure proper comments and details are added for further investigation.
Supervisors should periodically verify audit entries for accuracy.
Supervisor Responsibilities
Ensure all transactions for the shift are audited with no pending items.
Verify that executives categorize discrepancies accurately.
Randomly review audited transactions to check for accuracy and completeness.
Address any missing or unclear footage issues with the technical team.
By following this structured workflow procedure, the Heinen’s Checkout Task ensures accurate auditing of retail transactions, reducing errors and improving data integrity. Let me know if you need any refinements! 🚀
Last updated
Was this helpful?