Responsibilities
- Design, validate and productionize AI applications for parts manufacturing, process analysis and intelligent quoting.
- Process and analyze manufacturing data including CAD drawings, part features, process parameters, historical orders, machining records and quality results.
- Help build manufacturing knowledge bases, rule systems and reusable knowledge and tool capabilities from process expertise.
- Build drawing understanding, manufacturing knowledge retrieval, process assistance and intelligent quoting with large language models, RAG, multimodal models and Agents.
- Design prompts, context, tool calls, structured outputs and multi-step Agent workflows, including result validation, fallback handling and human confirmation.
- Use PyTorch, Transformers and related tools for model experiments, inference optimization, evaluation and application iteration, including fine-tuning or dataset construction when needed.
- Work with backend, product, project, process-engineering and customer teams on AI-service APIs, deployment, integration testing and production validation.
- Establish data-quality, model-effectiveness, Agent-task-success and issue-traceback mechanisms so that business users can understand and reliably use AI capabilities.
- Document experiments, evaluation sets, prompt standards, tool protocols and reusable components to keep AI applications maintainable and extensible.
Requirements
- Bachelor degree or above in computer science, artificial intelligence, mathematics, mechanical engineering, automation or a related field.
- At least three years of experience in deep learning, generative AI, Agent or industrial AI application development, including independent ownership of an AI capability through delivery.
- Strong Python skills and practical experience with PyTorch, Transformers, NumPy, Pandas or similar tools.
- Practical understanding of large-language-model applications, including model APIs, context design, RAG, embeddings, tool calls and structured outputs.
- Understanding of Agent workflow design, including task decomposition, state management, result validation, fallback handling and human confirmation.
- Strong problem-decomposition skills and the ability to turn business questions into testable AI tasks, evaluation sets and measurable outcomes.
- A strong engineering mindset, with attention to data quality, inference performance, API collaboration, cost control and production reliability.
- Industrial AI, smart manufacturing, machining, process planning, intelligent quoting or manufacturing-software experience is preferred.
- Experience with CAD parsing, geometric feature recognition, 3D-model processing, point clouds, computer vision or multimodal applications is preferred.
- Experience with model fine-tuning, inference deployment, evaluation systems, MCP, Tool Servers, knowledge graphs or multi-Agent collaboration is preferred.
- A demonstrable Agent, RAG or large-language-model application, evaluation report, technical blog or code repository is preferred; able to read English technical materials and communicate clearly across teams.
Role details are subject to final confirmation.