Core mission and direction
Responsible for the design, development and operation of native AI functions with search enhancement generation (rag), large language model (LLM), and artificial intelligence (AI Agent) as the core, in-depth study of product interface experience (UI/UX) and business processes to create a truly practical AI use experience; leading model selection, inference architecture construction, rag design and agent design combined with practical application scenarios, taking into account development quality and delivery efficiency; carrying out the entire process operation design covering effect evaluation, system monitoring, log design, and cost optimization, and continuously optimizing and creating a highly reliable AI use experience.
Specific job responsibilities
Design and implement the workflow of the large language model of the agent for the production environment with high stability; connect external services and internal application program interfaces (APIs) to build the running infrastructure of the agent; evaluate and filter the adapted external models, frameworks and services; clarify the actual needs of users through demand research, and collaboratively design the optimal agent solution across departments; build a test system and monitoring mechanism for the evaluation of the performance of the artificial agent to achieve the visual presentation of relevant indicators; establish a standardized development process and quality specifications through code review and document preparation; mine and solve various technical problems to ensure the accuracy, reliability, performance and scalability of the artificial intelligence system, and promote the continuous optimization of the system.
Job Requirements
Have more than 5 years of development experience using any language of Python, TypeScript, Go; have a large language model based interface (OpenAI/Anthropic/Google Gemini/Mistral, etc.); has practical work experience in artificial intelligence (LangChain/CrewAI/OpenAI Agent SDK/Google ADK/MCP, etc.) construction and contextual engineering; has experience in the design and development of search enhancement generation (rag), and is familiar with embedded technology, vector databases, and search algorithms; has experience in improving model generation accuracy through prompt word optimization and evaluation index design; has basic theoretical knowledge of natural language processing, and experience in data pipeline construction; has Japanese communication skills equivalent to Japanese proficiency test N1 level.
Priority
Has experience in the development of search infrastructure using tools such as Elasticsearch/OpenSearch/Vespa; has experience in system design that integrates knowledge graphs, structured data, and retrieval-enhanced generation (rag); has experience in system design and operation in cloud environments such as AWS/Google Cloud/Azure; and has experience in preprocessing and standardizing unstructured text such as customer emails, FAQs, and product documents.