Enhance Computer Vision Algorithms:

• Refine and optimize computer vision techniques to accurately detect common skin issues (e.g., acne, hyperpigmentation, wrinkles, dryness) using diverse, high-quality image datasets.

Develop a Basic Predictive Model:

• Create a straightforward predictive model to offer short-term skincare recommendations based on detected skin conditions.

Simplify Data Visualization:

• Build a basic interactive dashboard to present model insights, performance metrics, and actionable recommendations in an accessible format.
• 
  

We are looking for a  skilled AI/ML Developer to take our AI-powered Skin Analyzer to the next level.

Please check the reference website

https://getskinbeauty.com/ai/

Requirement Analysis & Project Proposal

• Tasks:
• Gather key requirements from Aurora Organics’ stakeholders.
• Define a focused project scope, success metrics, and KPIs tailored to computer vision and predictive modeling.
• Identify essential data sources (e.g., skin image datasets, basic product information).
• Deliverable:
• A concise project proposal outlining the focused objectives, methodologies, and timeline.

Data Collection, Cleaning & Exploratory Data Analysis (EDA)

• Tasks:
• Collect and aggregate skin image datasets and relevant product data.
• Perform data cleaning and preprocessing to ensure consistency.
• Conduct EDA to identify trends and initial insights related to skin conditions.
• Deliverable:
• A set of cleaned datasets

AI Model Development & Enhancement

• Tasks:
• Enhance computer vision algorithms to accurately detect skin issues.
• Develop and train a basic predictive model linking detected skin issues to short-term skincare recommendations.
• Validate the model using standard ML tools (e.g., Python).
• Deliverable:
• An improved AI model with performance metrics and detailed technical documentation.

Predictive Analytics & Data Visualization

• Tasks:
• Refine the predictive model based on feedback and testing outcomes.
• Build a simple interactive dashboard (using tools like Power BI or Tableau) to display model insights and performance.
• Deliverable:
• A basic predictive analytics model paired with an interactive visualization dashboard.