← back to jobs
> job detail
4
πŸ§ͺData Scientist

ML Performance Optimization Engineer

42dot Β· Pangyo, Gyeonggi-do, South Korea
// classified as
Data Scientist (Modeling, experiments, research.)
posted
1d ago
location
Pangyo, Gyeonggi-do, South Korea
languages
python, shell
tools
β€”
> stack
pythonshell
> description
About the Team & Mission

AD Division의 ML Performance Optimization EngineerλŠ” Autonomous Driving AI λͺ¨λΈμ΄ μ°¨λŸ‰ ν™˜κ²½μ—μ„œ μ•ˆμ •μ μ΄κ³  효율적으둜 λ™μž‘ν•  수 μžˆλ„λ‘ λͺ¨λΈ μ΅œμ ν™” 및 μ‹œμŠ€ν…œ μ„±λŠ₯ κ°œμ„  업무λ₯Ό μˆ˜ν–‰ν•©λ‹ˆλ‹€. Autonomous Driving 에 μ‚¬μš© λ˜λŠ” λ‹€μ–‘ν•œ λͺ¨λΈμ— λŒ€ν•΄ GPU/NPU 기반 μ΅œμ ν™”λ₯Ό μˆ˜ν–‰ν•˜λ©°, μ°¨λŸ‰ ν™˜κ²½μ—μ„œμ˜ λͺ¨λΈ 배포 및 μ„±λŠ₯ ν–₯상을 μœ„ν•œ 핡심 역할을 λ‹΄λ‹Ήν•©λ‹ˆλ‹€. λ˜ν•œ AI Model 및 Software νŒ€κ³Ό κΈ΄λ°€νžˆ ν˜‘μ—…ν•˜μ—¬ λͺ¨λΈ μ„±λŠ₯ 검증, μ‹œμŠ€ν…œ ν”„λ‘œνŒŒμΌλ§, μ„±λŠ₯ 뢄석 μžλ™ν™”λ₯Ό μˆ˜ν–‰ν•˜λ©° Autonomous Driving μ‹œμŠ€ν…œμ˜ μ•ˆμ •μ„±κ³Ό 효율 ν–₯상에 κΈ°μ—¬ν•©λ‹ˆλ‹€.

The ML Performance Optimization Engineer in the AD Division is responsible for optimizing AI models and improving system performance to ensure autonomous driving models operate efficiently and reliably in vehicle environments. This role focuses on GPU/NPU-based optimization for various autonomous driving models. You will play a key role in enabling seamless deployment and performance improvements on vehicle platforms. The role also involves close collaboration with AI model and software teams to conduct model validation, system profiling, and performance analysis automation for autonomous driving systems.

Responsibilities

  • GPU/NPU 기반 λ”₯λŸ¬λ‹ λͺ¨λΈ μ΅œμ ν™” μˆ˜ν–‰

  • λ”₯λŸ¬λ‹ λͺ¨λΈμ˜ μ„±λŠ₯ 검증 및 μ‹œμŠ€ν…œ ν”„λ‘œνŒŒμΌλ§ μˆ˜ν–‰

  • CPU, GPU, Neural Network Accelerator μ„±λŠ₯ 뢄석 및 μ΅œμ ν™”

  • μ‹œμŠ€ν…œ μ„±λŠ₯ 뢄석 툴 및 μ„±λŠ₯ 평가 μ§€ν‘œ 개발 μžλ™ν™”

  • μ°¨λŸ‰ ν™˜κ²½μ—μ„œμ˜ AI λͺ¨λΈ λ™μž‘ μ΅œμ ν™” 및 배포 지원

  • Optimize deep learning models using GPU/NPU acceleration

  • Validate deep learning model performance and conduct system profiling

  • Analyze and optimize CPU, GPU, and neural network accelerator performance

  • Develop and automate system performance analysis tools and evaluation metrics

  • Support deployment and runtime optimization of AI models in vehicle environments

Qualifications

  • 인곡지λŠ₯, λ¨Έμ‹ λŸ¬λ‹/λ”₯λŸ¬λ‹ κ΄€λ ¨ λΆ„μ•Ό 석사 ν•™μœ„ 이상 λ˜λŠ” 이에 μ€€ν•˜λŠ” κ²½λ ₯ 보유자

  • λ¨Έμ‹ λŸ¬λ‹ 및 λ”₯λŸ¬λ‹μ— λŒ€ν•œ 이해

  • CPU, GPU, NPU 기반 Low-level μ„±λŠ₯ μ΅œμ ν™” κ²½ν—˜

  • C/C++, Python, Shell Script 기반 개발 κ²½ν—˜

  • Linux, QNX, RTOS ν™˜κ²½μ—μ„œμ˜ 개발 κ²½ν—˜

  • μ‹œμŠ€ν…œ μ„±λŠ₯ 뢄석 및 Profiling κ²½ν—˜

  • Master’s degree or higher in Artificial Intelligence, Machine Learning, Deep Learning, or related fields, or equivalent practical experience

  • Strong understanding of machine learning and deep learning systems

  • Experience with low-level performance optimization for CPU, GPU, and NPU environments

  • Proficiency in C/C++, Python, and shell scripting

  • Development experience in Linux, QNX, or RTOS environments

  • Experience with system profiling and performance analysis

Preferred Qualifications

  • λ¨Έμ‹ λŸ¬λ‹ Workload μ΅œμ ν™” κ²½ν—˜

  • λ”₯λŸ¬λ‹ 기반 μ„ ν˜•λŒ€μˆ˜ μ—°μ‚° μ΅œμ ν™” 및 뢄석 κ²½ν—˜

  • Image Processing, Computer Vision, Robotics Algorithm μ΅œμ ν™” κ²½ν—˜

  • CUDA, MKL, SIMD, NEON 기반 μ΅œμ ν™” κ²½ν—˜

  • NVIDIA 기반 μ°¨λŸ‰ ν”Œλž«νΌ ν™˜κ²½μ—μ„œμ˜ 개발 λ˜λŠ” μ΅œμ ν™” κ²½ν—˜

  • Autonomous Driving κ΄€λ ¨ AI λͺ¨λΈ μ΅œμ ν™” κ²½ν—˜

  • Experience optimizing machine learning workloads

  • Experience optimizing and analyzing linear algebra routines for deep learning systems

  • Experience optimizing image processing, computer vision, or robotics algorithms

  • Experience with CUDA, MKL, SIMD, or NEON optimization techniques

  • Experience developing or optimizing systems on NVIDIA-based vehicle platforms

  • Experience optimizing AI models for autonomous driving applications

Interview Process

  1. μ„œλ₯˜ μ „ν˜•

  2. 1μ°¨ λ©΄μ ‘ (화상, 1μ‹œκ°„ λ‚΄μ™Έ)

  3. 2μ°¨ λ©΄μ ‘ (λŒ€λ©΄ ν˜Ήμ€ 화상, 3μ‹œκ°„ λ‚΄μ™Έ)

  4. 처우 ν˜‘μ˜Β·μž…μ‚¬

  1. Application Screening

  2. First Interview (Virtual, approximately 1 hour)

  3. Second Interview (In-person or Virtual, approximately 3 hours)

  4. Offer Discussion / Onboarding

Β 

Additional Information

  • μ „ν˜• μ ˆμ°¨λŠ” 일정 및 μ§„ν–‰ 상황에 따라 일뢀 변경될 수 있으며, 각 μ „ν˜• κ²°κ³ΌλŠ” λ“±λ‘ν•˜μ‹  μ΄λ©”μΌλ‘œ κ°œλ³„ μ•ˆλ‚΄λ“œλ¦½λ‹ˆλ‹€.

  • μ§€μ›μ„œ 제좜 μ‹œ μ£Όλ―Όλ“±λ‘λ²ˆν˜Έ, 가쑱관계, 혼인 μ—¬λΆ€, 연봉, 사진, 신체쑰건, μΆœμ‹  μ§€μ—­ λ“± μ±„μš©μ ˆμ°¨λ²•μƒ μš”κ΅¬ κΈˆμ§€λœ μ •λ³΄λŠ” μ œμ™Έ λΆ€νƒλ“œλ¦½λ‹ˆλ‹€.

  • μ§€μ›μ„œ μ ‘μˆ˜ 쀑 였λ₯˜κ°€ λ°œμƒν•˜κ±°λ‚˜ 기타 문의 사항이 μžˆμ„ 경우, recruit@42dot.ai둜 λ¬Έμ˜ν•΄ μ£Όμ‹œκΈ° λ°”λžλ‹ˆλ‹€.

  • κ΅­κ°€λ³΄ν›ˆλŒ€μƒμž 및 μ·¨μ—…λ³΄ν˜Έ λŒ€μƒμžλŠ” 관계법령에 따라 μš°λŒ€ν•©λ‹ˆλ‹€.

  • μž₯애인 고용 촉진 및 μ§μ—…μž¬ν™œλ²•μ— 따라 μž₯애인 등둝증 μ†Œμ§€μžλ₯Ό μš°λŒ€ν•©λ‹ˆλ‹€.

  • 42dot은 μ˜λ’°ν•˜μ§€ μ•Šμ€ μ„œμΉ˜νŽŒμ˜ 이λ ₯μ„œλ₯Ό λ°›μ§€ μ•ŠμœΌλ©°, μš”μ²­ν•˜μ§€ μ•Šμ€ 이λ ₯μ„œμ— λŒ€ν•΄ 수수료λ₯Ό μ§€λΆˆν•˜μ§€ μ•ŠμŠ΅λ‹ˆλ‹€.

  • μ§€μ›μ„œ λ‚΄μš© 쀑 ν—ˆμœ„ 사싀이 발견될 경우, μž…μ‚¬κ°€ μ·¨μ†Œλ  수 μžˆμŠ΅λ‹ˆλ‹€.

  • 인터뷰 ν”„λ‘œμ„ΈμŠ€ μ’…λ£Œ ν›„ μ§€μ›μžμ˜ λ™μ˜ν•˜μ— ν‰νŒμ‘°νšŒκ°€ 진행될 수 μžˆμŠ΅λ‹ˆλ‹€.

  • 3κ°œμ›”μ˜ μˆ˜μŠ΅κΈ°κ°„μ΄ 적용될 수 μžˆμŠ΅λ‹ˆλ‹€.

  • The recruitment process may change depending on schedule and progress; the result of each stage will be sent individually to your registered email.

    Please do not include legally prohibited information in your application (e.g., ID number, family relations, marital status, salary, photo, physical details, hometown).

  • For application errors or inquiries, contact recruit@42dot.ai.

  • Veterans and applicants eligible for employment protection will receive preferential consideration in accordance with applicable laws and regulations.

  • In compliance with the Act on Employment Promotion and Vocational Rehabilitation for Persons with Disabilities, registered individuals with disabilities will receive preferential consideration.

  • 42dot does not accept unsolicited resumes from search firms. We will not pay any fees for resumes submitted without prior agreement.

  • False information in your application may result in offer cancellation.

    A reference check may be conducted after the interview process, with your consent.

  • A 3-month probationary period may apply.