Some careers have more impact than others. If you’re looking for a career where you can make a real impression, join HSBC and discover how valued you’ll be. We are currently seeking an experienced professional to
About Payoneer Founded in 2005, Payoneer is the global financial platform that removes friction from doing business across borders, with a mission to connect the world’s underserved businesses to a rising global economy. We’re a community
ABOUT US Xsolla is a global commerce company with robust tools and services to help developers solve the inherent challenges of the video game industry. From indie to AAA, companies partner with Xsolla to help them
Some careers have more impact than others. If you’re looking for a career where you can make a real impression, join HSBC and discover how valued you’ll be. We are currently seeking an experienced professional to
AI院-GLM后训练团队-推理优化/推理Infra工程师(RL) 北京 全职 互联网 / 电子 / 网游 - 研发 职位描述 1. 基于SGLang进行大模型推理引擎与服务侧开发,提升吞吐、降低时延与成本,保障线上稳定性。2. 支持RL训练相关的推理链路,包括rollout生成、打分/奖励模型推理等,优化训练闭环效率与资源利用。3. 建设推理侧的监控与性能分析能力,参与线上问题定位与故障处理。4. 与算法/训练团队协作,将模型与训练需求落地到可运行、可维护的推理系统,并输出必要的技术文档。 职位要求 1. 熟练使用Python/C++(至少一种),具备良好的工程能力与性能意识。2. 熟悉GPU推理与常见深度学习框架(如 PyTorch),了解分布式通信与并行基本概念(NCCL 等)。3. 了解并使用过SGLang(或能快速上手并承担核心模块开发与调优)。4. 理解RL训练中推理侧的作用,能够与训练流程协同推进落地。5. 具备线上系统排障能力,能通过日志/指标定位性能与稳定性问题。加分项6. 有大模型推理系统、RLHF/RL 推理链路或训练-推理协同经验。7. 有推理性能profiling、算子/运行时优化经验。8. 有MoE模型优化经验。 投递...
Typical Accountabilities - Lead the development and execution of brand strategies, marketing plans, and promotional programs, to grow market share, improve customer engagement, and strengthen brand equity. - Analyse market and competitor data, oversee research initiatives,
Job Req Number: 121482 Time Type: Full Time Senior IT Specialist - SAP S/4HANA FICO consultant Purpose of the job: We offer you the opportunity to use your skills in highly complex tasks and to contribute
Purpose of Position The QA Engineer will be responsible for establishing, maintaining, and continuously improving quality assurance frameworks, testing processes, and quality standards across CASETiFYs digital products. The incumbent will work closely with product, engineering, design,
Tätigkeitsbereich:Forschung & Entwicklung incl. Design Fachabteilung:MB.OS Content Management and Rollout Gesellschaft:Mercedes-Benz Group China Ltd. Standort:Mercedes-Benz Group China Ltd., Beijing Startdatum:01.09.2026 Veröffentlichungsdatum:22.07.2026 Stellennummer:MER000469V Arbeitszeit:Vollzeit Bewerben Aufgaben Objective of job Mobile App and OTA Program Management role in MBGC
Tätigkeitsbereich:Forschung & Entwicklung incl. Design Fachabteilung:Software Lifecycle Management & Rollout Gesellschaft:Mercedes-Benz Digital Tech Ltd. Standort:Shanghai, Xingchi Tower, No. 399, Keqiao Road, Jinqiao, Pudong Startdatum:01.08.2026 Veröffentlichungsdatum:22.07.2026 Stellennummer:MER00046BT Arbeitszeit:Vollzeit Bewerben Aufgaben We are seeking an innovative, technology-driven, and benchmark-oriented
Agen沙盒工程师(环境平台方向) 深圳 社招 全职 技术 职位 ID:A215944 职位描述 负责构建和维护可扩展的多场景 Agent 沙盒环境,并建设 统一托管与调度平台。主要职责:1. 构建和维护可扩展的 Agent 沙盒环境,覆盖代码执行、命令行、浏览器/GUI、隔离 workspace 等形态,支撑评估、测试及后续业务扩展。2. 建设 Agent 沙盒统一托管与调度平台,提供创建、调度、回收与基础监管能力。3. 与 Agent 框架及业务方对接,沉淀环境模板、API 与使用规范,推动沙盒能力在团队内落地。 职位要求 1. 本科及以上学历,2 年以上 Agent 或 LLM 应用相关研发经验,做过需要执行环境 / 工具调用的 Agent 系统。2.
SCHOOL OF CONTINUING EDUCATION The School of Continuing Education, one of the seven Faculties/Schools of the University and a major provider of continuing and professional education in Hong Kong, offers a wide range of full-time and
26届校招-AI院-GLM后训练团队-推理优化/推理Infra工程师(RL) 北京 正式 互联网 / 电子 / 网游 - 研发 职位描述 职位描述1. 基于SGLang进行大模型推理引擎与服务侧开发,提升吞吐、降低时延与成本,保障线上稳定性。2. 支持RL训练相关的推理链路,包括rollout生成、打分/奖励模型推理等,优化训练闭环效率与资源利用。3. 建设推理侧的监控与性能分析能力,参与线上问题定位与故障处理。4. 与算法/训练团队协作,将模型与训练需求落地到可运行、可维护的推理系统,并输出必要的技术文档。 职位要求 职位要求1. 熟练使用Python/C++(至少一种),具备良好的工程能力与性能意识。2. 熟悉 GPU 推理与常见深度学习框架(如PyTorch),了解分布式通信与并行基本概念(NCCL 等)。3. 了解并使用过SGLang(或能快速上手并承担核心模块开发与调优)。4. 理解RL训练中推理侧的作用,能够与训练流程协同推进落地。5. 具备线上系统排障能力,能通过日志/指标定位性能与稳定性问题。加分项1. 有大模型推理系统、RLHF/RL 推理链路或训练-推理协同经验。2. 有推理性能profiling、算子/运行时优化经验。3. 有MoE模型优化经验。 投递...
About the team: Fanatics Collectibles is a new company working at the intersection of collectibles and technology. Backed by some of the best investors, sports properties, and operators in the world, we’re integrating with Fanatics’ global
Company Description H&M is a fashion brand that offers the latest styles and inspiration, from fashion pieces and unique designer collaborations to affordable wardrobe essentials. Our business idea is fashion & quality at the best price
Vantive is a vital organ therapy company on a mission to extend lives and expand possibilities for patients and care teams everywhere. For 70 years, our team has driven meaningful innovations in kidney care. As we
Tätigkeitsbereich:Forschung & Entwicklung incl. Design Fachabteilung:Software Lifecycle Management & Rollout Gesellschaft:Mercedes-Benz Digital Tech Ltd. Standort:Shanghai, 939 Hu Qing Ping Road, Min Hang District Startdatum:01.10.2026 Veröffentlichungsdatum:27.07.2026 Stellennummer:MER00046B8 Arbeitszeit:Vollzeit Bewerben Aufgaben Lead Software team of MB. OS OTA Solutions
At Johnson & Johnson,we believe health is everything. Our strength in healthcare innovation empowers us to build aworld where complex diseases are prevented, treated, and cured,where treatments are smarter and less invasive, andsolutions are personal.Through our
具身大模型训练框架工程师-觅蜂子公司 上海 校招 正式 技术族 - 算法类 职位 ID:A83406 职位描述 可以选择以下一个或多个方向深入参与:建设具身智能统一训练框架,支撑 VLA、世界模型、分层决策模型等具身基础模型的预训练与后训练,覆盖从十亿级到百亿 / 千亿级参数规模,并面向 300B+ 模型持续演进。建设大规模分布式训练系统,包括数据并行、张量并行、流水线并行、混合并行、MoE、长上下文、低精度训练与推理等,支撑万卡 GPU 集群上的稳定训练。建设规模化后训练与强化学习系统,涵盖模仿学习、在线 / 离线 RL、真机 RL,以及云端训练 + 边缘异步 rollout 的云边协同架构。建设训练效率优化体系,围绕数据、计算、异步、通信全链路进行度量与优化,包括 IO、算子融合、CUDA Graph、计算通信 overlap、梯度同步、性能 profiling 等。与算法团队协作,快速复现并工程化落地 VLA、世界模型、flow matching、RL 等前沿算法。 职位要求 计算机基础扎实,学习能力强,愿意深入复杂系统和底层细节。代码能力强,代码品味好,能熟练使用 AI
实习-AI应用开发工程师 合肥 校招 实习 数字技术 - 算法 硕士及以上 蔚来2027届实习生招募 职位 ID:A83495 职位描述 The Forward-Deployed AI Engineer serves as the primary technical interface between the Manufacturing AI Team and the production environment. The role is responsible for