Overview
We are looking for an experienced AI/ML Platform Integration Engineer to work on the integration and enablement of AI/ML accelerator runtimes within embedded Linux platforms.
In this role, you will work at the intersection of silicon, embedded software and AI, owning the integration of AI/ML runtime and inference components into Linux BSPs across multiple SoC platforms. You will collaborate closely with kernel/platform, silicon, runtime and application teams, while driving low-level bring-up, debugging and cross-platform enablement.
This is a hands-on engineering role requiring strong expertise in Yocto/OpenEmbedded, embedded Linux, C programming and hardware/software debugging, combined with practical exposure to AI/ML inference runtimes and hardware accelerators such as NPUs, GPUs or DSPs.
The role follows a hybrid working model and requires 3 days/week presence at the office in Bucharest.
Responsibilities
- Integrate AI/ML accelerator runtimes and inference libraries into embedded Linux BSPs;
- Develop and maintain Yocto/OpenEmbedded recipes, BitBake configurations, package groups and software layers;
- Drive platform bring-up and integration of accelerator support across multiple SoC/BSP targets;
- Collaborate with kernel/platform, silicon, runtime and application teams to coordinate releases and resolve integration issues;
- Investigate complex cross-layer issues involving the Linux BSP, drivers, runtime components and underlying hardware;
- Debug the complete host-to-accelerator data path for NPU, GPU and DSP-based solutions;
- Perform low-level hardware and board debugging using tools such as serial console, dmesg, gdb and Lauterbach;
- Analyze issues, identify the responsible software or hardware layer and provide clear, evidence-based technical findings;
- Support integration with external silicon and runtime vendors and provide reproducible defect reports;
- Enable a shared and modular runtime architecture across multiple platforms, minimizing platform-specific implementations;
- Contribute to improving integration processes, platform stability and overall maintainability of the AI/ML software stack.
Requirements
- 2+ years of hands-on experience with Yocto/OpenEmbedded, including BitBake, layers, classes and packagegroups;
- Strong programming skills in C, including native and kernel-level development;
- Good scripting skills with Shell and Python;
- Comfortable working with C++ runtime libraries;
- Hands-on experience with AI/ML inference runtimes and hardware accelerators, such as NPU, GPU or DSP;
- Strong understanding of multi-core application SoCs, including CPU clusters, interconnect/NoC architecture, memory maps and address decoding;
- Good understanding of clocks, resets, power domains and cache coherency;
- Practical experience with accelerator and peripheral bring-up and low-level debugging;
- Proven ability to debug the complete host-to-accelerator data path;
- Experience with embedded Linux BSPs, kernel/platform integration and hardware/software interfaces;
- Hands-on experience with hardware and board debugging tools, including serial console, dmesg, gdb and Lauterbach;
- Strong troubleshooting and root-cause analysis skills across software and hardware layers;
- Ability to collaborate effectively with cross-functional and distributed engineering teams;
- Good communication skills in English.
What We Offer
- Opportunities to grow through involvement in complex international projects;
- A collaborative environment with knowledge-sharing initiatives and technical communities;
- Meal vouchers as part of your monthly benefits;
- Private medical insurance, supporting your health and wellbeing;
- 25 days of annual paid leave under a full-time employment contract;
- Flexible working arrangements, aligned with project requirements and client environments;
- Access to technical training programs and professional development resources.
Ready to help bring the next generation of semiconductor and electronic products from concept to production? Join us and play a key role in delivering world-class quality, reliability and manufacturing excellence.
We welcome applications from both permanent employees and experienced freelance consultants. The engagement model will be discussed during the recruitment process and will depend on project requirements and candidate preference.
Please note that only suitable candidates will be contacted and that by applying to this role you are being informed about your personal data being processed by Avel Technologies.
You can find more details in our Privacy Policy: https://aveltechnologies.com/privacy-policy/