Introduction:
In today’s health-conscious environment, consumers are demanding transparency—not
just in price but in nutritional data. Grocery retailers like FairPrice Singapore host thousands
of SKUs with nutrition labels, yet this valuable information is not easily accessible or
structured for analytics.
Actowiz Solutions leveraged its AI-powered web scraping and data extraction engine
to automatically capture, clean, and analyze nutritional information from FairPrice’s product
listings. This allowed food brands, health startups, and analytics platforms to build
nutrient-rich databases for insights, labeling compliance, and dietary analysis.
Actowiz’s Scraping + AI Solution
1. Targeted Web Scraping Pipeline
Actowiz’s scrapers navigated FairPrice’s product pages and captured:
- Product title and category
- Ingredient list
- Nutrition facts table (when available)
- Serving size, calories, carbs, protein, sugar, sodium, etc.
- Brand and country of origin
2. Sample Extracted Data (Raw)
| Product |
Serving Size |
Calories |
Sugar (g) |
Sodium (mg) |
Protein (g) |
Category |
| Meiji Fresh Milk |
250 ml |
≈130 |
~12 |
≈120 |
≈8 |
Dairy |
| Kellogg’s Corn Flakes |
30 g |
≈108 |
~2 |
≈285 |
≈2 |
Cereal |
| Lay’s Classic Chips |
28 g |
160 |
1 |
170 |
2 |
Snacks |
3. AI Models Used
NLP Parsing – Extracted nutrition data from free-text and
OCR-type content
Unit Normalization Engine – Converted different nutrition
reporting units into 100g/ml base
Named Entity Recognition (NER) – Detected ingredient-level
allergens and additives
Data Clustering – Grouped products by nutrition profile (e.g.,
low-carb, high-protein)
Missing Data Imputation – Estimated nutrition values where labels
were incomplete
City-Wise SKU Monitoring in Singapore
Actowiz tracked availability and nutritional labeling from:
- Online listings across FairPrice’s Singapore-wide delivery zones
- Specific regional availability tags (e.g., “Available in Tampines, not in
Woodlands”)
- Store-only items (tagged but excluded from online delivery)
Real-World Insights Delivered
- Ready-to-eat meals had 2x sodium levels vs recommended daily intake
- Gluten-free labeling lacked consistency across cereals and snacks
- Frozen foods often missed full nutritional labels—Actowiz flagged 12+ brands
- “Healthy” categories still contained >20g sugar in popular juice SKUs
Testimonial
“Actowiz gave us instant visibility into nutrition data we previously couldn’t
access. It powers our app’s healthy recommendations in Singapore daily.”
– Founder, Singapore-Based NutritionTech Startup
Next Steps:
- Integrate scraped nutrition data with barcode/UPC databases
- Launch multi-platform scraping (e.g., RedMart, Cold Storage, Shopee Supermarket)
- Offer personalized meal plans based on scraped SKUs
- Train AI models to predict missing nutrition data using deep learning
Conclusion
Nutrition transparency is no longer a luxury—it’s a necessity. For platforms, food brands, and health-focused apps, access to structured, real-time nutritional data unlocks better product decisions and healthier consumer journeys.
Actowiz Solutions’ scraping and AI pipeline now powers nutrition intelligence at scale—starting with FairPrice Singapore and expanding globally.
