Overview
The building data blocks of CosmetiQ-MCP
Data Overview
CosmetiQ-MCP knits al least 5 data dimensions into a single, queryable dataset.
Each dimension is valuable on its own; together they unlock compound insights no silo could surface.
1 · Product Categories
We start by mapping every SKU to a precise category tree:
Make-up – Face, Eye, Lip, Cheek
Skincare – Face Treatments, Moisturisers, Eye Care, SPF, Wellness-Body, Cleansers, Masks, Lip Care
Haircare – Treatments, Styling, Shampoo & Conditioner
Fragrance – Women’s, Men’s, Unisex
2 · Ingredients
For every formula we link:
INCI-level ingredient list
Regulatory facts – functions, max concentrations, IFRA allergen limits
Featured-ingredient flags for quick filtering.
3 · Regions
Every product record can be replicated across different regions (e.g., France, Germany, Italy).
This lets you track things like local prices & promo cadence or region-specific sentiment.
4 · Consumer Profile
From review text and marketplace metadata we derive:
Gender tag – feminine, masculine, unisex
Age tag – teen, adult, mature
Skin type – dry, oily, combination, normal, sensitive
Skin concerns – acne, dullness, pigmentation…
5 · Consumer Feedback
We ingest consumer reviews in their original language (English, Spanish, and more).
Every review flows through Gen-AI pipelines that extract:
Cosmetic Effects
Effect reduced_acne, Effect increased_hydration
Frangrance Intelligence
scent_family, scent_intensity, pleasantness, off_notes
Texture Analysis
absorption_speed, afterfeel, weight
Usage & Rituals
use_context[], avoid_context[], step_order
Packaging Perception
material kudos, sustainability mentions
Value for Money
value-for-money score vs price tier
Because every data shares common keys (e.g. connecting consumer feedback to ingredients and skin profiles) you can hop from one dimension to another in a single query.
Last updated