【JAPAN AI】Product Manager, AI SaaS / English

株式会社ジーニー

給与：1200万円〜2000万円

雇用形態：正社員

勤務地：東京都

仕事内容

・Product Requirements & Architecture Design
‐Systematize product requirements from business goals and customer challenges, and translate them into technical architecture
‐Define feature boundaries, domain design, data contracts, and API specifications
‐Operate RFC-driven design review and decision-making processes
‐Design products leveraging LLM / generative AI (RAG, tool-use, guardrails, etc.)
・Technology Selection & Trade-off Decisions
‐Build vs Buy vs Integrate technology selection
‐Design non-functional requirements (availability, scalability, security, cost) and make quality attribute trade-off decisions
‐Formulate roadmaps for visualizing and systematically resolving technical debt
・Cross-functional Collaboration & Roadmap
‐Cross-functional collaboration with engineering teams, PdMs, and designers
‐Formulate and drive product roadmaps
‐Visualize technical debt and plan systematic resolution

Example Scenarios
The following are illustrative scenarios for this role:
◆Scenario 1: New Product Line Launch
Discover strong demand from enterprise clients for "AI agent-powered internal knowledge search." After market research, customer interviews, and technical validation, define product requirements. Lead RAG architecture design, data pipeline technology selection, and MVP scope definition, achieving beta release in 3 months.
◆Scenario 2: Build vs Buy Architecture Decision
When expanding voice AI capabilities, a decision is needed between in-house development vs external API integration. Conduct comparative analysis across 4 axes — cost, quality, development speed, and customizability — and file an RFC. After design review with the engineering team, decide on a hybrid approach (core functionality in-house, peripheral functionality via external APIs).
◆Scenario 3: Quality Attribute Trade-off Decision
JAPAN AI AGENT's P95 latency exceeds the target value. Analyze trade-offs between availability, cost, and latency, and propose introducing inference caching and optimizing model routing. Collaborate with the engineering team to implement, improving latency by 30% while keeping cost increase within 5%.

[Key Results (KR/Metrics)]
・Product roadmap achievement rate
・Rework rate attributable to architecture reviews
・New feature release cycle reduction
・Product quality attributes

[Tech Stack]
・Languages: Python (backend), TypeScript / React / Next.js (frontend) / NX
・AI/LLM: LangChain, LangGraph, JAPAN AI STUDIO SDK, RAG, Agent Framework
・Infrastructure: GCP (containers / K8s), Docker, Terraform
・Tools: Slack, Confluence, Linear, Google Workspace, GitHub, Notion
・AI Dev Support: Claude Code MAX Plan, Cursor, ChatGPT, Devin
・Hardware: Mac (Apple Silicon), dual monitors

応募資格

・5+ years of practical experience as a Product Manager or Tech Lead
・Practical experience as a software engineer (at a level where you can understand the architecture of an entire product)
・Experience designing products that leverage LLM / generative AI (RAG, tool-use, guardrails, etc.)
・Deep understanding of system design (distributed systems, microservices, event-driven architectures)
・Experience in requirements definition and consensus-building with business stakeholders
・Language requirement (at least one):
・Japanese: Fluent — able to discuss product development without friction
・English: Business level or above and Japanese proficiency (JLPT N2 or equivalent), or a minimum of one year of experience working in a Japanese-speaking environment.