腾讯CoDesign-客户端开发工程师

We’re on a mission to make financial services accessible and effortless for SMEs. As one of Singapore’s latest digital wholesale banks fully regulated by MAS, we’re committed to continuous innovation to bring about simpler, safer and more rewarding financial services. #bringingaboutwhatsnext We are looking for Product Commercialization Specialist to join us. Job Overview Your task as Product Commercialization Specialist is to increase the company's success and profitability with of our existing products. The spectrum of your tasks ranges from innovative product innovations, to optimising the efficiency of existing products through data-driven analysis of product performance by product type, channels, markets; and ensuring optimized utilization, pricing and revenue generated by our products. You will own the repository & governance of the product programs, and contribute to the design of new product programs by including details of market research, pricing strategy and potential profitability of the product. You will also own the repository of new product approvals, consolidating the details and approval trails for audit purposes. You will also be tasked to conduct market research and product viability studies, to ensure our products addresses the pain points of the target segment & market. In addition to this, you will develop sustainable strategies in order to build on the success in close co-operation with all departments, through co-designed product campaigns, campaign performance analysis so as to provide insights on improving the profitability of our product suites. Job Responsibilities Work with Business Units and conduct market feasibility studies, covering target segment pain points, competitor analysis, and pricing & profitability studies. Work with Product Development (PD Tech) and conduct product design & market feasibility prior to Product Requirements Design stage; and conduct pricing analysis, implementation impact analysis to support implementation and go-to-market strategy. Collaborate with Business Units and Business Intelligence to design data-driven campaigns, campaign performance analysis and A/B testing. Track and monitor product performance across channels, markets & customer profiles, and develop insights on persona of customers and maximize process & profitability based on targeted customer segments. Own repository for product programs and product approvals; ensuring the versions are up to date and the appropriate approvals and renewals are conducted timely. Draft & co-design new product programs & new product approvals, by consolidating and reviewing inputs from Business, Risk (Credit, Operational, Fraud), Legal & Compliance and Operations.

蚂蚁DeepXPU团队专注在基于GPU等AI芯片的高性能异构计算系统研发，欢迎加入我们一起致力于： ● Software&Hardware Co-Design：深入分析GPU等AI芯片的硬件架构，打磨极致高效的算子与显存优化技术，优化大模型引擎，构建异构虚拟化能力。通过系统建设和优化，推动计算效率逼近硬件的性能极限，打造面向大模型的AI系统生态，提升智算集群的整体效率。 ● Algorithm&System Co-Design：从系统的视角出发，和算法团队紧密配合，共同探索大模型结构和范式的创新，并面向大规模AI应用场景进行系统级优化与落地。

我们是淘天集团阿里妈妈工程平台团队，致力于打造面向搜推广场景及GenAI与AI Agent应用的AI Infra平台。团队服务于淘天电商广告核心业务，承担 广告模型超大规模Embedding表征学习和Sparse-Dense模型及Dense大模型（LLMs/Diffusion等）训练与推理，AI Infra研发和优化等关键职责，在高性能在线服务&训练平台（推荐系统/多模态大模型/大语言模型）、分布式系统（计算/存储/网络）、异构计算和AI编译优化（GPGPU/CPU）等课题上都具有业界前沿的挑战。 1. 负责超大规模Sparse-Dense模型（for 广告）和Dense大模型（for GenAI应用）的训练架构设计优化并推动生产落地。 2. 负责大规模异构硬件Training系统的计算性能与效率优化。 3. Algorithm-Training-Inference Co-Design，整体提升模型效率。

参与快手大语言/多模态基座模型、业务应用的研发，工作内容包括： 1、深度参与多模态/大语言模型训练全链路开发，包括数据、预训练、后训练全流程优化; 2、设计和优化分布式训练框架，通过混合并行，通信计算overlap、低精度训练等方法解决超长序列、超大规模moe场景下的训练效率问题； 3、参与通用高性能RL框架的开发和优化； 4、算法工程co-design，探索最优的训练范式。