C++/CUDA Developer

DataRoot Labs is a full-cycle AI R&D center and the largest talent and compute pool in Ukraine. In 8+ years in business, we’ve completed over 45 AI-enriched projects for numerous industry leaders and startups in Automotive, Energy and natural resources, Navigation, Gaming, Education, and more.

Our client provides autonomous trucks for logistics yard operations, improving safety and efficiency by replacing human drivers.

You’ll be responsible for making the AI stack drive the vehicle properly!
 

Key Responsibilities:

• Write and maintain efficient and scalable code in C++ with a strong focus on CUDA
• Profile and fine-tune CUDA kernels for performance, memory usage, and scalability
• Work with multi-threaded and parallel computing concepts to maximize GPU throughput
• Work closely with ML engineers, system architects, and DevOps to deliver performant systems
• Provide input on software design, architecture, and best practices in GPU-accelerated environments
• Design performance tests and benchmarks to validate software under real-world loads
   

Requirements:

• 4+ years with C++, including modern standards (C++11/14/17)
• 1+ year of hands-on experience with CUDA development
• Strong understanding of parallel computing and GPU optimization techniques
• Familiarity with profiling tools such as Nsight, Valgrind, or CUDA Profiler
• Good understanding of memory management, multithreading, and data structures
• Intermediate English (written & spoken) for documentation and collaboration.
   

Would be a plus:

• Experience in Automotive
• Knowledge of machine learning or computer vision pipelines
   

What We Offer:

• Startup culture, a strong goal-oriented team, and a research mindset
• Opportunity to leverage your engineering skills for fellow engineers and shape the future of AI
• Working with the newest technical equipment
• 20 working days of annual vacation leave
• English courses, Educational Events & Conferences
• Medical insurance