Backend Engineer, AI Products

This full-time backend engineering role develops and operates APIs for JAPAN AI STUDIO and other in-house AI products. Responsibilities include adding product features, improving performance, refactoring code, and maintaining regular weekly releases. The engineer will also participate in design reviews and technology selection, with decision-making responsibility based on delivered output.

The role includes implementing LLM and AI-agent capabilities in products. Work covers prompt design, tool-call control, and construction of RAG pipelines in coordination with frontend engineers and a Product Architect. Other responsibilities include designing and implementing API integrations with customer SaaS services such as Salesforce, SmartHR, and kintone, as well as data synchronization processes.

Applicants need at least three years of professional software engineering experience and must have either backend or API development experience with a modern language such as Python, TypeScript, Go, Kotlin, or Rust; development experience using AI, generative AI, LLMs, or AI coding tools; or development and operations experience in AWS, GCP, or Azure. Native-level Japanese is required. Experience in Python or Go backend development, TypeScript, React, Next.js, LLM or AI-agent applications, Docker, Kubernetes, Terraform, RDBMS design or tuning, large-scale data processing, external API integrations, or data integration platforms is welcomed. Specific years of experience with a particular technology stack are not required, and learning Python and TypeScript after joining is acceptable.

The development environment includes Python for backend work; TypeScript, React, Next.js, and NX for frontend work; LangChain, LangGraph, and the JAPAN AI STUDIO SDK for AI and LLM development; GCP with containers and Kubernetes; Docker; BigQuery; PostgreSQL; and customer data sources. Slack, Confluence, Linear, Google Workspace, GitHub, Notion AI, Claude Code MAX Plan, Cursor, ChatGPT, and Devin are also listed as tools. Mac computers with Apple Silicon and dual-monitor support are provided.

The position is based in Shinjuku, Tokyo, with hybrid work consisting of three office days and two remote days per week. Working hours are 10:00–19:00, with Saturdays, Sundays, and public holidays off. The employment is permanent full-time employment with a one-month probationary period. The posting lists annual compensation of JPY 7,000,000–14,000,000. The body also lists monthly compensation of JPY 500,000–1,000,000, including base pay of JPY 359,551–719,102 and fixed overtime pay for 45 hours of JPY 140,449–280,898. Overtime beyond 50 hours is paid separately, with salary review opportunities twice a year, bonuses twice a year, and a stock-option grant system.

Benefits include company-paid AI tools, support for paid development tools up to JPY 30,000 per year, book-purchase support up to JPY 30,000 per half-year, Japanese or English learning and certification support, a refresh allowance up to JPY 5,000 per month, housing support up to JPY 30,000 per month for eligible areas, club activity support, periodic lunch and dinner support, refresh leave after three years of continuous employment, annual health checks, employee stock ownership, complete social insurance, and full transportation reimbursement. The selection process is document screening, a coding test, four to five interviews, and an offer; a reference check is required before the final interview.