Customising Librosa for Unique Audio Processing

I'm seeking an expert in AI and audio processing who can help me tailor the Librosa library to meet specific needs. I want to adjust how we handle audio data, refine our feature extraction parameters, and modify file handling methods.

Key Tasks:
- Customizing audio processing algorithms: Specifically, I need assistance with time-stretching, pitch-shifting, beat tracking, and sound track separation from a music file.
- Tweaking feature extraction parameters: The Mel-frequency cepstral coefficients (MFCC), Chromagram, and Spectral contrast need to be tailored to suit this project.
- Customizing file handling methods: I want to change how the library interacts with our audio files.

Ideal skill set includes:
- Extensive experience with the Librosa or Similar for Audio Separation library
- Background in AI
- Proficiency in audio processing
- Ability to customize libraries to fit unique project parameters.

Please note, the goal is to utilize our own parameters and change the library's response accordingly. This project requires a deep understanding of audio processing and the ability to creatively adapt the tools at hand.