(Senior) AI/ML Engineer -Embedded/Automotive Performance Optimization (f/m/d)

Position Description

We are collaborating with a renowned staffing service provider in Belgrade, Serbia, Looking Glass D.O.O. As part of this partnership, we are actively seeking talented individuals for various positions, which will be locally employed by our Serbian partner company.
Join and be a part of a dynamic and innovative team, working under a contractual agreement with our esteemed Serbian staffing service provider.

As a (Senior) Software Engineer at Cinemo, you will play a critical role in the development and enhancement of our AI/ML powered applications for wide range of automotive hardware and OS platforms such as Android Automotive OS (AAOS) and Linux. Your primary responsibility will be to design, develop, and maintain high-quality Cinemo’s in-vehicle infotainment software solutions that meet the needs of our customers.

In this role, you will:

• Focus on ensuring that state-of-the-art models run with maximum performance on wide range of automotive platforms
• Make the most out of embedded/automotive hardware and software platforms in the field of AI/ML
• Elevate the in-car experience across different car-lines worldwide by crafting the best-of-class AI-powered infotainment functionality
• Optimize and tune the best ML models for different HW/SW platforms
• Ensure that every bit of available performance is used for the best possible user experience, including optimization to different CPU/GPU architectures and usage of dedicated hardware IP blocks available on the platform

What you will need to succeed:

• Experience in the areas of NLP, Machine Learning, Deep Learning, Data Preprocessing
• Experience with optimization of performance-critical algorithms for different CPU (ARM / x86) and GPU architectures (Nvidia)
• Experience with state-of-the-art performance toolkits, software stacks and profiling tools (Cuda / CuDNN / TensorRT, Jit + Jax)
• Proficient knowledge of Python, C/C++, ARM/x86 assembly, PyTorch, TensorFlow, Keras
• Fundamental knowledge about parameter efficient fine-tuning techniques (e.g., LoRA, PEFT)
• Good written and verbal English communication skills