Job Description
Responsible for the research and development of basic cloud computing and big data technologies, including but not limited to the following directions:
- Cloud infrastructure technology, including developing networks (such as RDMA, programmable chips) for million-level servers, heterogeneous computing servers, and data centers, and building intelligent operation and maintenance systems for large-scale infrastructure (such as AIOps)
- Virtualization technology, including improvements to open-source technologies like XEN and KVM, as well as our self-developed SDN, VPC, and other network virtualization and storage virtualization technologies. Also includes lightweight container solutions like Docker.
- Improvements to open-source database kernels such as MySQL, PostgreSQL, MongoDB, Redis, and HBase
- Includes large-scale distributed storage systems (such as file systems, KV stores, BigTable, etc.) containing tens of thousands of nodes in a single cluster, multi-site and multi-cluster distributed computing systems (MapReduce, DAG, MPI parallel computing, Batch, Hive/Spark-like computing systems including offline, distributed development languages, distributed development IDEs, query optimization, stream real-time computing, graph computing, MPP, etc.), elastic distributed resource management and scheduling (massive multi-dimensional multi-target scheduling systems, multi-resource dimension resource isolation technologies, etc.), machine learning platforms (including Paratemter Server, deep learning, logistic regression, etc.), and computation on new hardware types such as heterogeneous computing (including CPU, GPU, FPGA, RDMA, etc.)
- The goal of the big data online engine system is to integrate advertising, search, and recommendation into one, achieving intelligence from the service-side to the mobile-side in cycles of nearly 100 milliseconds. It supports overall models of nearly TB level and completes various recalls such as knowledge reasoning vector matching, and other deep learning sorting and prediction algorithms. Billions of commodities participating in the calculation are kept up-to-date in real-time, supporting hundreds of algorithm engineers to conduct tests in numerous scenarios. Under the abstract of the operator flowchart, the models and data inside the engine can be adjusted and laid out at any time to meet the needs of iteration.
Position Requirement
- Quick learner, constantly breaking through technical bottlenecks, eager to explore unknown fields, and ready to face new challenges at any time
- Possess the basic qualities of an excellent programmer, have a cleanliness obsession with code, can dig deep into problems, utilize tools effectively, and have insights into engineering efficiency.
- Full-stack talent, with expertise in computer system architecture, operating system kernels, networking, distributed systems, algorithms, etc., and able to integrate them
- At least a mastery level in one programming language, not limited to one development language, and have the willingness and ability to continuously learn multiple development languages