My SkinIQ repository is named skiniq.
The repo is owned by @aditiraj0129-netizen.
The primary language of the codebase is Python.
Project work started in 2026-08 and was last updated in 2026-08.
A live demo is available at https://www.linkedin.com/posts/aditi-raj-330459295_hi-im-going-to-show-you-a-project-i-built-ugcPost-7491060975115108352-b8i8/?utm_source=share&utm_medium=member_desktop&rcm=ACoAAEdvk5QBYfJ9UV8Vn-FCLUDidVIViS9HHKY.
The ingredient gazetteer contains over 28,000 INCI names from the EU CosIng database.
A curated risk table covers roughly 80 ingredients.
The SQLite schema has seven or more tables, ready for Postgres migration.
I built six specialized agents—Vision, OCR + Ingredient, Coordinator, Profile Resolver, Recommendation, and Symptom—each using techniques like conformal prediction, EasyOCR, and rule‑based reasoning to keep the system honest and explainable.
python -m app.db.seed_ingredients
python -m app.db.seed_gazetteer
python -m app.db.seed_products
python -m app.db.seed_symptom_knowledge