Optivize - Quick Start Technical Guide

Getting up to speed on what we built and how it works

🤖 FEATURE 1: AI Product Predictor

“Will my product succeed?” – Let AI decide

🛠️ What it Does

Users can train an AI model to predict if their product will be successful. They upload data about past products—price, marketing budget, how widely it’s sold, etc.—and get a score from 0–100 indicating the likelihood of success.

🧠 Technical Implementation

  • Frontend Architecture: Event-driven JavaScript with debounced input validation and progressive form enhancement
  • Data Flow:
    User Input → JSON Validation → API POST → ML Pipeline → Feature Engineering → Model Training/Inference → Response Rendering
  • ML Model: RandomForestRegressor (Scikit-learn) with feature importance calculation and cross-validation
  • API Design:
    • POST /api/train: 
      {
  "samples": [
    {
      "product_type": "Laptop",
      "seasonality": "Fall",
      "price": 899.99,
      "marketing": 8,
      "distribution_channels": 9,
      "success_score": 87
    }
  ]
}
    • POST /api/predict: 
      {
  "product_type": "Laptop",
  "seasonality": "Fall",
  "price": 899.99,
  "marketing": 8,
  "distribution_channels": 9
}
    • GET /api/history: Returns paginated prediction history with optional filters

💡 Smart Features

  • Feature Engineering: Categorical encoding for product types, numerical scaling for price ranges, and temporal encoding for seasonality patterns
  • Real-time Insights: Post-prediction analytics such as:
    • Price positioning vs. market averages
    • Marketing effectiveness score
    • Seasonality pattern analysis

✨ How to Use the AI Product Predictor

1. Prepare Your Training Data

Before making accurate predictions, train your AI model with historical product data that includes known outcomes.

Required fields for training:

  • product_type: e.g., Laptop, Sneakers
  • seasonality: Winter, Holiday, Year-Round, etc.
  • price: Product price in USD
  • marketing: 1–10 scale for marketing reach
  • distribution_channels: 1–10 scale for availability
  • success_score: Outcome score (0–100) based on real-world performance

Submit the data to: ```http POST /api/train