Senior Python Audio Engineer (DSP & AI Music)

About Cryo Mix

Cryo Mix is an AI-powered online mixing and mastering platform that empowers artists of all genres to achieve professional-quality audio production. Our platform performs in-depth analysis of tracks and emulates the processes of top-level mixing engineers, handling everything from vocal processing to mastering.

What We Do

After conducting extensive analysis of vocal and instrumental files, our AI identifies resonances caused by imperfect recording environments, determines optimal compressor settings, makes micro EQ adjustments across more than 1000 frequency bands, de-esses vocals, modifies instrumental EQ to accommodate vocals, and much more.

Our Workflow

• Upload - Artists drag & drop stereo mixes or individual stems directly in their browser.

• AI Processing - Cryo Mix balances levels, enhances vocals, and cleans up tone with real-time progress tracking.

• Interactive Mixing - Users can choose presets, chat with their mix in plain language, or dive into advanced settings.

• Preview & Export - Listen to before/after comparisons and export high-quality masters (WAV).

The Opportunity

We are looking for a Python Audio DSP Expert to architect and optimize the core audio processing engine that powers Cryo Mix. While our frontend ensures a smooth user experience, you will be responsible for "the ear" and "the hands" of the operation—ensuring that the audio output sounds indistinguishable from a human engineer's work.

This is a contract role focused on improving our Python-based signal processing pipeline. You will translate musical concepts (warmth, punch, clarity) into mathematical reality using Python, ensuring efficient processing of high-fidelity audio at scale.

Key Responsibilities

• Algorithm Development & DSP

  • Design and implement high-quality audio effects chains (EQ, Multiband Compression, Saturation, Limiting) in Python.

  • Translate AI analysis data into precise DSP parameters to fix resonances, sibilance, and masking issues.

  • Develop algorithms for "Vocal Riding" and automatic gain staging based on genre specific targets.

• Audio Analysis

  • Improve our feature extraction pipeline (BPM detection, Key detection, RMS/LUFS metering, Spectral analysis).

  • Optimize the detection of specific audio artifacts (clipping, background noise, plosives).

• Performance & Infrastructure

  • Optimize Python audio processing pipelines for speed and memory efficiency (reducing cost per mix).

  • Manage handling of large audio files (stems) using effective buffering and streaming techniques.

  • Ensure lossless input/output quality across various file formats (WAV, AIFF, MP3).

Technical Stack

You will be working with a specialized audio stack:

• Core Language: Python 3.10+

• Audio Libraries: Librosa, NumPy, SciPy, SoundFile, Pydub

• DSP Frameworks: TorchAudio or custom DSP implementations.

• Processing: FFmpeg for transcoding and file manipulation.

• Architecture: Redis/Celery for job queues, communicating with a Laravel backend.

• AI/ML: PyTorch/TensorFlow (for integrating inference models into the pipeline).

Specific Focus Areas

We need focused attention on:

• Vocal Chain Perfection: Fine-tuning the pitch detection and correction algorithms to sound natural, not robotic.

• Loudness Maximization: Implementing transparent limiting that achieves competitive commercial loudness (LUFS) without introducing audible distortion.

• Stem Mixing Logic: Improving the algorithm that balances the relationship between the vocal stem and the instrumental stem (frequency masking/ducking).

• Processing Speed: Refactoring code to reduce the time it takes to process a full song from minutes to seconds.

Ideal Candidate Profile

• Required Experience

  • 5+ years of Python development with a heavy focus on audio/math.

  • Deep understanding of Digital Signal Processing (FFT, Filter design, Convolutions).

  • Musical background: You understand what a compressor feels like, not just how it works mathematically. You know the difference between "muddy" and "warm."

  • Experience analyzing audio data (Spectral centroids, MFCCs, zero-crossing rate).

• Technical Expertise

  • Expertise in NumPy/SciPy for vectorization.

  • Familiarity with FFmpeg command line and bindings.

  • Experience building scalable backend workers (Celery, RQ, or similar).

What Success Looks Like

• Our AI masters sound competitively loud with retained dynamics.

• Vocal processing is free of sibilance artifacts and sits perfectly in the pocket of the mix.

• The Python worker processes tracks 30% faster than current benchmarks.

• You have established a robust testing pipeline that prevents audio regressions.

Why Work With Us

• Work on a cutting-edge AI audio processing platform used by artists worldwide.

• Tackle the "Holy Grail" of audio engineering: automating creativity.

• Collaborate with a team passionate about empowering creators.

• Flexible remote work