どんな仕事かAIによる生成

AI Engineerとして事業部のエンジニアリングチームに入り、社内モデルを用いたエージェントの実装や評価、プロンプトの改善を行います。プロトタイプから本番環境への展開と運用を担い、サードパーティAPIから社内モデルへの移行も支援します。

応募資格・募集要項・仕事内容を見る

応募資格

必須

3 to 6 years of professional software engineering experience, including hands-on work building LLM applications.
Practical agentic engineering experience: you have built agents using tool and function calling or an orchestration framework and have debugged them when they misbehave.
Working knowledge of prompt engineering: structured outputs, iteration against test cases, and the understanding that prompts need versioning and measurement.
Strong fundamentals in at least one modern language (e.g., Python, Java, Go), and comfort with APIs, backend services, and testing.

歓迎

Japanese language ability - a significant plus. Our business unit teams work in Japanese day to day, and being able to work in the language directly makes the embedded model far more effective.
Retrieval-augmented generation experience.
Familiarity with evaluation frameworks or LLM observability tooling.
Exposure to self-hosted inference (e.g., vLLM) or cloud model deployment.

求める人物像

Curiosity about how models behave - you want to understand why an output was wrong, not just retry it.
Comfortable working embedded in someone else's engineering team, communicating directly with business stakeholders, and operating with a degree of ambiguity.

募集要項最終確認 11/12

職種: AIエンジニア
雇用形態: 記載なし
給与: 記載なし
勤務地: 東京都
働き方: 記載なし
経験年数: 3年以上
掲載日: 2026/09/02 18:19（72日前）
最終確認: 2026/11/12 21:35（1日前）

仕事内容

Department Overview

The LLMAD team drives adoption of Rakuten AI and open-source LLMs hosted within our own environment. We're hiring AI Engineers who can be embedded directly inside business units to drive real adoption of in-house large language models in place of third-party alternatives.

Position Details

As an AI Engineer you will be embedded inside a business unit's engineering team and build with them day to day: implementing agent workflows and integrations, iterating on prompts, running evaluations, and getting features live in production on our in-house models.
You will work on an engagement alongside a senior engineer who sets the overall agent architecture and model direction, and you will take growing ownership of the delivery as you build depth. This is hands-on applied LLM work, close to the product.
Embed inside a business unit's engineering team and implement agent and tool-use workflows for their use case - function calling, multi-step flows, retrieval, guardrails, and fallback handling - writing production code in their codebase.
Get features from prototype into production and support them there: debugging failures, tightening latency and cost, and fixing what breaks.
Run task-level evaluations of in-house and open-source models for your use case, and extend and maintain the eval sets and harnesses together with the business unit.
Design and iterate prompts, output schemas, and tool definitions - versioned and measured against evals rather than tuned by feel - and help narrow quality problems down to the prompt, the retrieval, the model, or the data.
Support migrations from third-party APIs to our in-house models, and surface the gaps you find to the senior engineer on the engagement and to our model and platform teams.
Document what worked as reusable recipes, examples, and starter templates for the team's shared library, and help business unit engineers adopt our tools and APIs.

この求人は、楽天グループ株式会社の採用ページの内容をAIが読み取って整理したものです。応募の前に採用ページの原文を確かめてください。データの集め方企業サイト