Engineer AI Application
Magna International Inc.10 hours ago
Nanchang, ChinaMid Level
Responsibilities
- Develop processes and governance for AI projects, model management, data labeling, and AI-system operations.
- Plan and architect enterprise AI applications and data-intelligence initiatives and advise on technology selection.
- Design and implement internal knowledge-base systems, including document parsing, chunking, vector retrieval, and semantic reranking.
- Develop LLM-based agents with function calling and integrations with internal systems such as MES and ERP.
- Evaluate, design, implement, train, optimize, and deploy production-line machine-vision defect-detection systems.
- Guide image-data collection and labeling, improve model accuracy, quantize and compress vision models, and deploy them reliably to edge devices such as industrial computers.
- Manage the full AI-project lifecycle, including requirements research, solution design, development, rollout training, ongoing maintenance, and technology evaluation.
- Follow cybersecurity, data-protection, AI-ethics, and company information-security requirements and identify system security risks.
Requirements
- Bachelor’s degree or higher in computer science, artificial intelligence, software engineering, automation, mathematics, or a related field, plus English proficiency at university CET-4 level or above.
- At least three years of relevant AI software-development experience.
- Deep understanding of neural networks, including CNN, RNN, and Transformer architectures, backpropagation, activation functions, optimizers, model selection, and tuning.
- Expertise in Python, Pandas, NumPy, Scikit-learn, and either PyTorch or TensorFlow, with familiarity with Linux and Shell scripting.
- Experience with RAG, vector databases such as Milvus or Qdrant, large-model fine-tuning using SFT or LoRA, prompt engineering, and intelligent-agent development using frameworks such as LangChain, LlamaIndex, or Dify.
- Experience with OpenCV and classic object-detection models such as Unet and YOLO, plus familiarity with model quantization and TensorRT edge deployment.
- Knowledge of SQL and relational databases including Oracle, HANA, MSSQL, MySQL, and PostgreSQL, as well as BI/data-warehouse integration and ETL processes.
- Ability to independently complete requirements analysis, data preparation, model selection, development, training, production deployment, documentation, and stakeholder communication.
- Strong problem-solving skills, initiative, and willingness to debug systems on-site in a production workshop.
- Manufacturing, especially automotive, AI-project implementation experience is preferred.
Benefits
- Regular, permanent employment with Magna Powertrain.
- Engaging and dynamic automotive-technology environment with varied responsibilities and career-development prospects.
- Work may include on-site debugging in production workshops and deployment to production-line edge devices.
- Magna is an equal-opportunity employer with a diverse-workforce commitment.
Tech Stack
Categories
About Magna International Inc.
Magna International is a public Canadian automotive supplier that designs and manufactures vehicle systems, modules, and electronics for global automakers, and also builds complete vehicles through Magna Steyr. Founded in 1957 and headquartered in Aurora, Ontario, it operates worldwide as a Tier-1 supplier across body, seating, powertrain, ADAS, and electrification. Its common shares trade on the TSX (MG) and NYSE (MGA).