Jr. / Mid Machine Learning Engineer – Time-Series & Inertial AI

Role Overview

Jr./Mid Machine Learning Engineer – Time-Series & Inertial AI at Si-Ware Systems in Riyadh, focused on deep learning applications for MEMS IMU sensors and inertial systems.

Role Purpose

Pioneer the use of deep learning to enhance the raw performance of MEMS IMU sensors through software, simulation, and data-driven modeling. Work with datasets to train models that correct stochastic errors, denoise signals, and improve sensor performance.

Key Responsibilities

Time-Series AI Development

• Design, train, and validate neural networks (CNNs, LSTMs, TCNs, Transformers) to denoise raw inertial sensor data.
• Fuse denoised sensor outputs for improved overall performance.

Virtual Sensing

• Develop AI models that enhance MEMS IMU sensor outputs using self-supervised or supervised learning techniques.

Data Pipeline Engineering

• Build robust data processing pipelines to handle massive, high-frequency inertial systems datasets.
• Implement filtering, normalization, augmentation, and windowing of inertial data.

Rapid Prototyping

• Implement ML algorithms based on published academic research papers.

Research

• Regularly survey state-of-the-art academic research papers at the intersection of machine learning and inertial systems.

Model Optimization

• Quantize and prune trained PyTorch/TensorFlow models for deployment on resource-constrained embedded targets (e.g., ARM Cortex-M).

Cross-Functional Collaboration

• Work closely with Egypt-based Systems and Firmware teams to ensure models are designed within the computational limits of edge microcontrollers (TinyML).

Skills & Competencies

• Deep learning frameworks: PyTorch and TensorFlow.
• Neural network architectures: CNNs, LSTMs, TCNs, Transformers.
• Time-series data modeling and signal processing.
• Model quantization and pruning techniques.
• Data pipeline design and implementation.
• Embedded machine learning (TinyML) and ARM Cortex-M constraints.