We are Unlock Design, a recruitment agency based in Japan, and the job description below is an opportunity with one of our forward-thinking client companies. We aim to ensure a trustworthy and considerate experience for our candidates, emphasizing their privacy and long-term career goals. See our full promise and approach by visiting https://ud-intl.com/pledge/.

【Basic Information】

Client ID: 106000

Job ID: 2411

Position: 生成AIを活用した大規模B2Cチャットアプリケーションプロジェクトの開発リーダー

Location: Tokyo

Salary: 8M ~ TBA JPY

Language: Japanese + English (Business)

Contract type: Permanent

Who we are

私達は、AI・ディープラーニング技術を活したソリューションにより、経営・社会課題の解決に挑むテックイノベーションファームです。特に、画像解析ディープラーニング、センサーによる異常検知AI、最適化AIなど様々なデータに対応するAIを組み合わせた「マルチモーダルAI」に強みを持ちます。戦略策定から要件定義フェーズに始まり、現場のコンサルテーションから開発・運保守まで、投資対効果を実感するまで気通貫で伴し、これまでにもごみ焼却運転の動化など多くの実績をあげています。また社会課題にも積極的に取り組んでおり、SDGs課題と環境変化を衛星画像から発する「GRASP EARTH」などで第4回、第5回宇宙開発利賞を連続受賞しました。 今後も技術とビジネスのみを追求し、社会・顧客が持続的に効果を実感できる最峰のソリューションを提供します。

事業内容

・AI・ディープラーニング技術のコンサルティングおよび開発

・共同事業、ライセンス、保守モデル、自社開発等によるプロダクトの提供

・人工衛星データを活用したAI解析ソリューションの提供

• English here WE are a tech innovation firm that addresses business and societal challenges using solutions powered by AI and deep learning technologies. We specialize in "Multi-Modal AI," which combines various AI solutions to handle diverse data, including image analysis deep learning, anomaly detection AI using sensors, and optimization AI. We are involved in every aspect of the process, from strategic planning to requirement definition, on-site consultation, development, and operational maintenance, ensuring a comprehensive approach that delivers a return on investment. We have a track record of numerous achievements, including automating waste incineration operations. Furthermore, we actively engage with societal challenges and have received consecutive awards, such as the 4th and 5th Space Development Utilization Awards, for our work in discovering SDGs issues and environmental changes from satellite imagery, including "GRASP EARTH." In the future, we will continue to pursue the heights of technology and business, providing the highest level of solutions that allow society and our clients to experience sustainable results. Business Activities:
  • Consulting and development of AI and deep learning technologies.
  • Providing products through joint ventures, licensing, maintenance models, and in-house development.
  • Offering AI analysis solutions utilizing artificial satellite data.

Roles and Responsibility

LLMベースのチャットボットサービスを開発・運用する大規模な専門家チームのリードとしてご活躍いただきます。技術的専門知識を活かし、多くのステークホルダーと連携しながら、チャット機能が国や企業レベルのポリシーを遵守することを保証し、デプロイされたサービスの継続的な品質向上を行っていただきます。

AWS上でのバックエンドクラウドアプリケーション開発及び、最新のLLMアプリケーション開発のベストプラクティスをビジネスの現場に活用する機会を提供できます。また、チームのシニアメンバーとして、ビジネス上の課題解決においてリーダーシップを発揮し、チームメンバーの品質担保も実施頂きます。

• English here As the lead of a large, specialized team developing and operating LLM-based chatbot services, you will utilize your technical expertise to ensure that chat functionalities adhere to national and corporate policies while collaborating with numerous stakeholders. You will also focus on the continuous improvement of deployed services. You will have the opportunity to apply best practices in backend cloud application development on AWS and in cutting-edge LLM application development to business scenarios. Additionally, as a senior team member, you will demonstrate leadership in solving business challenges and ensure quality assurance among team members.

Minimum Requirements

• 開発マネジメント経験
• LangchainおよびPythonを使用したLLMシステム開発の基礎
• RAGシステムの設計および実装
• LLMベンチマークの作成および管理
• AWSの基本知識およびサーバーレス設計および実装(例:ECS、ECR、Lambda、RDS、ElastiCache、ネットワーキングなど)
• Infrastructure as Code技術(Terraform)
• Pythonにおけるクリーンな非同期API開発の知識(FastAPI、Asyncio)
• AWSでのソフトウェア保守・運用知識(ログを用いたデバッグ、根本原因の分析、論理的問題およびパフォーマンスボトルネックの調査)
• アジャイル開発サイクルに関する知識(開発チケットの作成、分割および更新、優先順位付け、継続的なイテレーション)
• Git-flowおよびGitHub flowに関する知識

上記すべてを満たす必要はなく、上記に近しい経験があり、キャッチアップが可能であることが示すことができれば可

• English here
  • Experience in development management
  • Basics of LLM system development using Langchain and Python
  • Design and implementation of RAG systems
  • Creation and management of LLM benchmarks
  • Fundamental knowledge of AWS and serverless design and implementation (e.g., ECS, ECR, Lambda, RDS, ElastiCache, networking)
  • Infrastructure as Code skills (Terraform)
  • Knowledge of clean asynchronous API development in Python (FastAPI, Asyncio)
  • Software maintenance and operations knowledge on AWS (debugging with logs, root cause analysis, investigation of logical issues and performance bottlenecks)
  • Knowledge of agile development cycles (creating, splitting, updating, prioritizing development tickets, continuous iteration)
  • Knowledge of Git-flow and GitHub flow
  Meeting all of the above criteria is not mandatory. Experience close to the listed requirements and the ability to catch up are acceptable.

Preferred Requirements

• 最先端のLLMのアップデートに関する知識と継続的な学習意欲
• LLMエージェンティックシステムの設計および実装
• AWSのML/LLM関連サービス(SageMaker、Bedrock、OpenSearchベクトルストアなど)に関する知識・ノウハウ
• エンジニアの小規模チームを率いた経験
• LLMのファインチューニングに関するノウハウ(例:Hugging Faceを使用)
• デザイナーやプロダクトマネージャーなどの非エンジニアとの効果的なコミュニケーション
• プロダクトのモニタリングおよび可観測性ツール(Amplitude)
• B2C関連のユーザー実験、A/Bテストなど
• TypeScriptおよびSQLに関する知識
• 一般的なデータ分析・データサイエンススキル
• 英語の熟練度
• English here
  • Knowledge of cutting-edge LLM updates and a strong desire for continuous learning
  • Design and implementation of LLM agent-based systems
  • Expertise in AWS ML/LLM-related services (e.g., SageMaker, Bedrock, OpenSearch Vector Store)
  • Experience leading small engineering teams
  • Knowledge of fine-tuning LLMs (e.g., using Hugging Face)
  • Effective communication with non-engineers such as designers and product managers
  • Experience with product monitoring and observability tools (e.g., Amplitude)
  • Experience with B2C user experiments and A/B testing
  • Knowledge of TypeScript and SQL
  • General data analysis and data science skills
  • Proficiency in English

Conditions / Benefits

Employment type: Full time

Location: Chiyoda-ku, Tokyo

Annual Salary: 8M ~ TBA JPY (Negotiable)

・Salary is determined based on experience, and skills

・Annual salary system (1/12 of the annual salary is paid monthly) ・Above includes 45 hours of overtime (additional overtime will be paid separately)

・Salary review once a year (there are many cases of salary increases for high performance in just 6 months upon joining the company)

・Bonus twice a year (January and July)

Hours: 9:30 - 18:30 (standard working hours: 8 hours)

・General office hours: 9:30 ~ 18:30

・Flexible working hours during probation

・Core time: 10:30 - 16:30

・Flexible time: (start) 7:30 - 10:30 (end) 16:30 - 20:30

・Breaktime: 1 hour between 11:00 - 16:00

・Two days off per week (Saturdays and Sundays)

・National Holidays

・Year-end and New Year holidays (December 28 - January 4)

・Refreshment leave (5 days after 1 year of employment)

・Paid Holidays: 10 days of paid leave - amount depends on the years of employment

・Before annual paid leave is granted (6 months after joining the company), 3 days of special leave granted upon joining

Probation: 6 months

Benefits:

・All transportation expenses are paid

・Overtime allowance

・Casual clothing is welcome

・Telework consultation (frequency of attendance varies depending on the department and duties)

【Insurance】

• Employment insurance
• Workers accident compensation insurance
• Health insurance: Kanto IT Software Health Insurance Association membership
• Employee pension

Interview Process

1. CV Screening
2. Technical Test
3. 1st interview
4. 2nd Interview
5. Last Interview
6. Offer