Backend Developer for Mood-Driven Photo Editor

Role Overview

Join us in creating a cutting-edge mood-driven photo editor that transforms images based on users' emotional states, focusing on raw visual storytelling without filters or gimmicks. You will develop a backend Lambda function that accurately analyzes photos and returns structured segmentation data. This data will be pivotal for OpenAI to determine how each image region should be edited according to specific moods, such as "grateful," "melancholic," or "nostalgic."

Responsibilities

• Develop a scoped one-off AWS Lambda function using Python, OpenCV, and NumPy.
• Accept input in the form of an image (base64 or file) and a mood string for tagging.
• Output a JSON object containing:
  • Unique region IDs
  • Semantic labels like "highlight_zone" or "shadow_patch"
  • Bounding boxes and polygons for each region
  • Average Hue/Saturation/Lightness values
  • Percentage of the image covered by each region
  • Base64-encoded masks for regions matching the original image dimensions
  • Global image dimensions and image ID
• Ensure the function builds the emotional anatomy of photos to facilitate future edits by OpenAI.

Required Skills

• Proficiency in Python with a strong focus on OpenCV for image processing, including the handling of image masks and HSL or HSV color spaces.
• Experience with AWS Lambda and familiarity with deployment tools like Zappa or Serverless Framework.
• Ability to produce clean, well-documented Python code that returns structured and testable JSON outputs.

Nice to Have

• Understanding of facial segmentation and emotion/scene tone analysis.