AI Agent Application Development Engineer

This role develops production AI agent applications tailored to customers’ specific business operations, including customer support, internal knowledge search, and workflow automation. Responsibilities cover requirements hearing, understanding business processes and end-user needs, UI and UX design, frontend and backend implementation, API development, LLM integration, deployment, operational monitoring, usage analysis, and improvement proposals. The engineer is responsible for delivering applications that customers continue to use in their actual work.

Frontend work includes designing and implementing conversational interfaces, dashboards, and business screens with TypeScript, React, and Next.js. The role also covers LLM streaming displays using Server-Sent Events, real-time updates, and interaction design. Backend responsibilities include implementing LLM integration APIs, business logic, authentication and authorization, and external system integrations with Node.js and TypeScript. Mastra may be used in TypeScript environments, while LangChain and LangGraph may be used in Python environments.

Required qualifications include at least three years of professional web application development experience, experience across both frontend and backend development, application development and operations in at least one of AWS, Azure, or GCP, and experience designing and implementing web APIs using REST API or GraphQL. Candidates must also have experience developing applications using LLMs or generative AI, integrating external AI or ML services such as OpenAI API or Bedrock, or developing and operating user-facing web applications. Team development experience and understanding of Git, code review, and CI/CD are required.

Mastra experience is especially welcomed, although support is available for candidates with TypeScript implementation experience who have not used Mastra. Additional welcomed experience includes Next.js App Router and Server Components, LLM streaming, Server-Sent Events, WebSocket, LangChain, LangGraph, LlamaIndex, AI agent patterns such as ReAct, Tool Use, Multi-Agent and MCP, UI/UX design, design systems, Amazon Bedrock, Azure OpenAI, Vertex AI, and financial-sector compliance such as FISC guidelines.

The engineer works with AI solution architects and AI agent solution development engineers and may participate in technology selection and architecture design. Projects are organized by engagement rather than assignment to one customer and may span finance, manufacturing, satellite data, and infrastructure. The position is based in Otemachi, Chiyoda-ku, Tokyo. It is full-time and indefinite-term, with a six-month probation period. The work uses a discretionary labor system with eight deemed working hours per day; during probation, a flextime system applies. Remote work is available with, in principle, one office day per week. Annual compensation is JPY 7,000,000–13,000,000, including deemed overtime pay for 45 hours per month, with excess overtime paid separately. Bonuses are paid twice yearly based on performance. Benefits include social insurance, transportation allowance, paid leave, refresh leave, learning and career support, equipment and book expenses, healthcare support, and side-job eligibility subject to conditions.