Applications: candidates may apply from outside Japan. Visa sponsorship: available. Japanese: Business level. English: Business level. Japan Dev tags: Engineering, Web / Full-stack, AI.
About Rakuda AI
Rakuda AI is a Tokyo-based AI product company building agentic LLM systems, applied-AI products, and AI-native software for the Japanese market and beyond. We design, train, and ship full AI systems in-house — from agent architecture and retrieval pipelines to evaluation harnesses and model fine-tuning — entirely in-house, with no external subcontracting.
Our founders are serial entrepreneurs:
-
Masaki Arakawa(Representative Director): Keio University. Founder of Amy Inc., a software-development and marketing-support company, which he successfully exited through a sale to Vector Inc. (Tokyo Stock Exchange Prime-listed).
-
Yudai Futai (Co-founder): Kyoto University, Faculty of Economics. Former e-commerce consultant at Rakuten, Manager of Business Development at IoT startup Qrio, and co-founder of scouty (now LAPRAS Inc.), a leading Japanese AI HR-tech company.
We ship production-grade AI products on aggressive timelines — typically 1 to 1.5 months from kickoff to production. Recent work includes an agentic document-analysis system for a fintech (LLM + RAG + multi-step planning), AI feature integration for a B2B SaaS in three weeks, and a custom evaluation pipeline that lifted task success rate from 62% to 94% on a production agent.
About the Role
We are hiring our first Principal AI Engineer / Founding Engineer to define the AI engineering foundation of Rakuda AI as we scale into a category-defining AI product company. You will own the design of our agentic systems, LLM infrastructure, and applied-AI engineering practice end-to-end.
This is not a "manage a team of mid-level engineers" role. We are looking for someone who can personally architect, evaluate, and ship the hardest AI systems we build — agent harnesses, retrieval and reranking stacks, evaluation frameworks, fine-tuning and distillation pipelines — and who can set the technical bar for every AI engineer we hire after them. The successful candidate will work side-by-side with the founders, with full authority over model selection, architecture, hiring, and engineering culture.
We are targeting senior AI engineers currently at frontier AI labs (OpenAI, Anthropic, Google DeepMind, Mistral, Cohere, etc.), AI-native product companies (Cursor, Perplexity, Sakana AI, etc.), or Staff / Senior Staff / Principal-level AI specialists at top global or Japanese tech firms.
Responsibilities
- Own the technical architecture of our AI systems: agent harnesses, LLM orchestration, retrieval / RAG, evaluation infrastructure, fine-tuning and model distillation pipelines
- Personally implement the hardest parts of each system — novel agent architectures, evaluation methodology, inference-cost and latency optimization, prompt programs at scale
- Lead applied-AI engineering practice: model selection, prompt engineering, eval frameworks, AI safety and guardrails, MLOps
- Set engineering standards across AI and full-stack work: code review, architecture review, framework choices, infrastructure and security baselines
- Hire and develop the AI engineering team alongside the founders
- Work directly with founders and senior client stakeholders to shape AI product direction
Requirements
- 8+ years of professional software engineering experience, including at least 3 years at Staff / Senior Staff / Principal level (or equivalent) at a top-tier engineering organization, with substantial recent AI / LLM work
- Demonstrated track record of personally architecting and shipping production AI systems at scale (agents, RAG, evaluation, or fine-tuning)
- Deep applied-AI engineering: LLM orchestration, agent design, retrieval / RAG, evaluation methodology, prompt engineering, vector databases, fine-tuning (SFT / RLHF / DPO) or model distillation
- Strong full-stack proficiency as a shipping capability: TypeScript and at least one modern JS framework (Next.js, React, or equivalent), plus strong backend, database, and cloud infrastructure (AWS, GCP, or equivalent)
- Strong opinions on engineering quality and on what "production-grade AI" actually means
- Business-level English (for technical communication, AI tooling, and engineering documentation)
- Business-level Japanese (for direct communication with Japanese clients and executives) — required
- Based in Japan and able to work full-time on-site at our Tokyo office
Strongly Preferred
- Prior experience at a frontier AI lab (OpenAI, Anthropic, Google DeepMind, Mistral, Cohere, etc.) or an AI-native product company (Cursor, Perplexity, Sakana AI, etc.)
- Track record building production agent systems beyond single-shot prompting (planners, multi-step reasoning, tool use, memory)
- Authorship or major contributions to widely-used AI open-source projects (LangChain, LangGraph, DSPy, vLLM, etc.) or peer-reviewed AI publications at NeurIPS / ICML / ICLR / ACL
- Prior experience as a founding engineer, principal engineer, or technical co-founder at a startup
- Experience working directly with C-level stakeholders or as a technical decision-maker
What We Offer
- Compensation: starting from JPY 30M base, with significant upside tied to the growth of the company (negotiable based on level and contribution)
- Direct collaboration with serial-entrepreneur founders with prior exits and operational track records
- Genuine founding-engineer authority: model and architecture choices, hiring decisions, engineering culture
- AI-first engineering environment with full access to frontier AI tools (Claude, Cursor, OpenAI, internal tooling) and infrastructure budget for serious experimentation
- Tokyo office in the central business core of the city
- On-site, in-person work — designed for high-velocity decision-making and deep collaboration with founders, AI engineers, and clients
Working Style
This is an on-site role at our Tokyo office. We believe the best AI engineering work — particularly at the principal level, and particularly during this stage of the company — happens in the same room as the founders, the rest of the AI engineering team, and our clients. We do not offer remote work for this position.