AI Brand Detection & Excel Automation

Overview

CHHAT’s AI Brand Detection & Excel Automation system by neWwave — A specialized AI solution that identifies cigarette brands from retail shelf images and automatically converts the results into structured Excel-ready data.

Description

CHHAT conducts large-scale surveys across small shops in Cambodia to understand which cigarette brands are available at each store.

Each survey can include thousands of store records and close to 100,000 shelf images. Reviewing every image and entering the identified brands into Excel manually would require significant time, cost, and human resources.

neWwave developed an internal AI system that analyzes the submitted shelf images, identifies visible cigarette brands, connects the results to the relevant store records, and generates structured data for Excel-based reporting.

‍

The Challenge

Retail shelf images are captured in real-world conditions, where products may appear under different lighting, angles, reflections, shelf arrangements, and levels of visibility.

The client needed a faster and more consistent way to analyze a large volume of images without relying on workers to review and record every cigarette brand manually.

The Solution

neWwave created a proprietary computer vision and data-automation solution tailored to CHHAT’s survey process.

The client provides the store data and related shelf images. The system analyzes the images, identifies the cigarette brands, matches the results to the correct store records, and produces a structured Excel-ready output.

This transforms the process from:

Shelf images → Manual checking → Manual Excel entry

into:

Shelf images → AI brand detection → Automated Excel-ready results

‍

Our Role in Building the System

neWwave designed and developed the complete internal AI solution for CHHAT, covering automated brand recognition, high-volume image processing, store-level data matching, Excel automation, structured reporting, and quality evaluation. The system was built to process large survey datasets efficiently while keeping the underlying AI technology and infrastructure secure and proprietary.

Features We Engineered

• AI Cigarette Brand Detection
• Image-to-Excel Automation
• Large-Scale Image Processing
• Store-Level Brand Identification
• Confidence-Based Result Evaluation
• Structured Excel-Ready Reporting
• Reference Image Management
• Quality Review & Improvement Workflow
• Scalable Retail Intelligence System

‍

Performance & Business Impact

The optimized system can process up to approximately 24,000 store records and close to 100,000 retail shelf images in around three hours.

This replaces a process that would otherwise require workers to inspect the images and enter the detected cigarette brands into spreadsheets individually.

By automating the workflow, CHHAT can:

• Complete large-scale surveys significantly faster
• Reduce manual review and data-entry costs
• Apply consistent identification standards
• Connect detected brands to the correct store records
• Receive structured data ready for Excel analysis
• Scale future surveys without increasing the manual workload at the same rate

‍

Why This Project Matters to neWwave

The CHHAT AI Brand Detection & Excel Automation system demonstrates how specialized AI can transform a repetitive and expensive manual task into a fast, scalable operation.

Instead of using a general-purpose recognition tool, neWwave created a solution around the client’s specific survey process and reporting needs.

For neWwave, this project represents the kind of AI transformation we aim to deliver: measurable, secure, practical, and designed to reduce both operational time and cost.

‍