Background: Iris is dedicated to empowering content creators through innovative technology. We're looking for skilled Audio Software Engineers to help us develop cutting-edge solutions.

Task: Design and prototype an audio analysis system that can detect sound levels and other audio features from audio files, then serve this information through an API. This system should assist content creators, particularly video editors, in efficiently identifying and navigating audio events e.g. in their video content.

Guidelines:

1. Concept and Design:
   • Define the needs that your system could potentially address.
   • Outline the system's architecture and key features.
   • Describe the audio features your system will detect, for instance:
     1. Sound level variations (e.g., low vs. high sound levels)
     2. Specific sound events (e.g., laughter, car sounds, door slams, etc.)
2. Implementation:
   • Develop a basic prototype that demonstrates core functionalities.
   • Choose any programming language and audio processing libraries that suit you.
3. Workflow:
   • Illustrate how your system operates, from receiving audio input to providing analysis results via API.
   • Outline the API endpoints and response format for the analyzed data.
4. Presentation:
   • Prepare a brief explanation of your system, covering:
     1. Key design decisions and their rationale in the context of video editing
     2. Challenges faced and your solutions
     3. Potential future improvements and additional features

Submission:

• Concept description with explanations and any relevant diagrams
• Prototype implementation, including code for audio analysis and API
• A brief presentation during the tech interview

Note: Focus on showcasing your audio processing knowledge, problem-solving skills, and understanding of how audio analysis can streamline video editing workflows. The goal is to demonstrate your approach to designing audio systems that provide real value to content creators by helping them efficiently navigate and edit their video content based on audio cues.

We usually aim at having candidates complete the assignment in roughly one week, but if you can't make it in one week (due, for example, to work or other personal reasons), just tell us upfront and it won't be a problem. Feel free to ask us any questions you think of. Once you hand in your solution, we'll have a look at it and then we'll contact you back with decision regarding a next step in the process.

One more thing: please don't share neither the assignment nor your solution online or with anyone, so as to avoid having other candidates in the future get to know about it. We suggest either sending it as a zip file or cloud link.