神州数码高级AI开发工程师(3）(J20659)
任职要求
本科及以上学历，计算机、软件工程、人工智能等相关专业优先。 3-5年Python开发经验，具备扎实的编程基础和良好的编码习惯，至少掌握Flask/Django/FastAPI等其中一种主流Web框架。 熟练掌握AI应用开发相关技术，需掌握以下技能的4项及以上： 理解大模型提示词工程，有良好的提示词编写习惯，能对提示词调优。 理解并实践过RAG（检索增强生成）方案； 使用或集成过向量数据库（如FAISS、Milvus、Weaviate等）； 熟悉LangChain…
工作职责
参与公司AI应用项目的设计、开发与交付，负责核心功能模块的实现与优化。 深度参与基于大语言模型（LLM）的应用开发，包括RAG（检索增强生成）、Agent（智能体）等技术方案的落地。 负责AI项目中数据流、模型调用、检索框架等系统架构的搭建与性能优化。 跟进AI前沿技术，持续优化和升级现有AI应用，保障产品的稳定性与高效性。 跨团队协作，推动AI项目顺利交付，支持客户的定制化需求。
1、负责视觉、音频、影像、感知等算法的工程架构设计 2、负责算法在移动端、云端、PC端等多端的落地交付 3、负责算法的极致性能优化，包括CPU/GPU/NPU/DSP等 4、负责算法性能专项攻坚、疑难问题解决 5、负责AI模型部署领域、高性能计算领域新技术的探索研究
The Role TESLA is offering a full-time IT Support DevOps AI position in the Information Technology Department (Work Location: Tesla Giga Factory Shanghai). If you are a versatile expert integrating AI development, DevOps practices—someone who can efficiently tackle challenges, solve complex technical problems in user support and experience scenarios, and reject repetitive and inefficient work patterns—this role is perfect for you. IT Support DevOps AI is a core role connecting the company’s IT systems and user-facing processes, standing at the forefront of enhanced user support implementation. You will engage in work across multiple domains, including AI technology R&D, containerized deployment, and operational support. Through technical practice, you will support the company in optimizing user interactions, improving support efficiency, and contributing to the core goal of user experience transformation. Responsibilities • Undertake AI algorithm R&D, model optimization, and training, with a strong emphasis on fine-tuning (FT), supervised fine-tuning (SFT), reinforcement learning (RL), and advanced tuning techniques; focus on user support scenarios such as data analysis, query resolution, issue detection, and automated assistance to ensure AI technology aligns with user experience needs. • Complete the deployment, monitoring, and scaling of AI solutions based on container technologies like Kubernetes (K8s) and Docker, ensuring high availability and stability of the system in the operational environment, while integrating AI underlying technologies like neural networks and Transformer architectures for efficient performance. • Participate in DevOps process development, optimize the full lifecycle of AI model and system development, testing, and deployment, and realize automated deployment, continuous integration (CI), and continuous delivery (CD), incorporating RL-based optimization and model tuning for adaptive user support systems. • Collaborate with user support-related departments such as helpdesk, customer service, and product teams to deeply understand user pain points and provide data-driven AI technical solutions, leveraging SFT and attention mechanisms to enhance personalized user experiences. • Respond quickly to technical requirements and faults in user-facing systems, troubleshoot issues in AI systems, container clusters, and network environments, minimize impacts on user interactions, and improve support efficiency and satisfaction through advanced AI tuning and underlying model diagnostics. • Track cutting-edge technologies in the AI and DevOps fields (e.g., large language models with FT/SFT/RL integration, cloud-native operations) and industry trends, promote the pre-research and application of new technologies in user support scenarios, and continuously optimize system performance using techniques like model compression and quantization.
1. 主导大模型系统架构设计： 负责RAG系统的整体架构设计，包括存储层、检索层、推理层与缓存层的技术选型与实现。 构建高可用、低延迟的分布式推理服务框架，支持向量数据库集成（如Milvus、Elastic）、知识库管理与多模态检索优化。 设计并实现Agent工作流编排框架，支持工具调用（MCP协议）、任务规划与自动化执行。 2.模型部署与性能优化： 负责大模型（LLM/VLM）的本地化部署、量化压缩、动态批处理与推理加速，优化GPU/CPU异构算力利用率。 3. AI服务平台开发： 基于Java/Go/Python构建高并发、可扩展的AI微服务，与现有业务系统深度集成，实现模型训练-部署-监控的全链路管理。 4.技术领导与跨团队协作： 指导中级工程师，制定技术方案，并主导技术攻关。与产品、算法、基础设施团队协作，定义需求并推动工程落地。
1、负责剪映CapCut稳定性方向Agent功能模块的开发并稳定支撑剪映系业务，助力传统工程转向Agent化、全球化； 2、基于业务使用场景，将Agent能力持续深度赋能到业务中； 3、通过AI能力建设起全方位数据运营体系，包括但不限于可用性指标、历史事故、资源利用率等，发现系统薄弱点，推动落地并持续优化。