AI Software Engineer, New Graduate (In-Vehicle Log Analysis)

This role is ideal for an MSc or PhD student with a strong background in AI/ML model development. You will conduct research to evaluate and benchmark various model architectures for "low-training" or few-shot learning applications specifically for embedded vehicle logs.

This a temporary role that runs for approximately 12 weeks with a projected start date in early June.

• Research & Benchmarking: Conduct an AI research study to identify efficient neural architectures for high-accuracy log analysis.
• Model Development: Develop and evaluate few-shot learning applications for embedded vehicle logs (e.g., ADB, DLT).
• Experimental Validation: Perform experimental validation of models in resource-constrained environments using internal compute unit data streams.
• Documentation: Clearly document findings, experiments, and results of various benchmarked architectures.

• Must have graduated in Spring 2026 with a BS, MS, or PhD degree with a focus on AI/ML model development.
• AI/ML Expertise: Strong understanding of few-shot learning, neural architecture search, or anomaly detection.
• Technical Proficiency: Strong programming skills in Python and deep learning frameworks like PyTorch or TensorFlow.
• Prior experience with NLP and language models (e.g., SBERT).
• Must be local to the bay area, or willing to relocate, and come into office 3 days/week (Monday, Tuesday and Thursday).

• Experience with "low-training" applications in resource-constrained or embedded environments.
• Familiarity with automotive logs such as ADB or DLT.

Sonatus Perks include:

• Complimentary lunches, snacks and beverages during on-site working days
• Hybrid office work arrangement
• Computer accessory allowance
• 15 paid holidays