## Project Overview We are looking for a Senior Machine Learning Engineer to optimize several existent voice machine learning models. These models analyze human voice and generate inference from it. The optimization project includes the following tasks: ### Key Responsibilities 1. **Retraining a Model (PyTorch)** - Enhance the current Audio models by retraining them using PyTorch to improve performance and accuracy. 2. **Exporting as an ONNX** - Convert the optimized PyTorch models to the ONNX format for easier deployment and integration. 3. **Achieve Over 70% Accuracy** - Ensure the optimized models achieve a minimum accuracy of 70% in generating soft skills from voice data. 4. **Label Data** - Organize and label existing data to improve the quality of the training datasets. 5. **Get New Datasets** - Source and integrate new datasets to further enhance the model's performance and accuracy. 6. **Require Label Effort in an Effective Manner** - Implement efficient labeling strategies to maximize the effectiveness of the dataset. ### Skills and Qualifications - **Audio Signal Processing (MUST)**: Strong understanding and experience working with PCM (pulse-code modulation) and DSP (digital signal processing). - **Strong Proficiency in PyTorch**: Extensive experience in developing, training, and optimizing machine learning models using PyTorch. - **Data Management**: Expertise in organizing, labeling, and creating large datasets. - **Experience with ONNX**: Demonstrated ability to export and debug ONNX models. ### Good to Have - **Rust knowledge**: We are looking to optimize our internal systems using Rust. - **Previous work with Soft Skills**: Experience in analyzing human soft skills is a plus. - **Server Development**: Proficiency in developing servers using FastAPI or Rust with Axum.