【Ai Workforce】AI/LLM Division Research Engineer (R&D)

Job Description

#### Background

The Ai Workforce Division promotes the adoption of our generative AI platform, **“Ai Workforce,”** among leading enterprises in Japan, contributing to significant improvements in corporate productivity. These enterprises rely daily on a wide variety of business documents, including contracts, financial statements, purchase orders, invoices, proposals, and more.

At Ai Workforce, we develop LLM-powered workflow engines and AI agents to automate our customers’ complex business processes.

As the number of customers and projects continues to grow, we are required to develop, execute, and evaluate workflows and AI agents with increasingly high standards of quality and efficiency. The Senior R&D Research Engineer in the Ai Workforce Division aims to practically solve the challenges we face by effectively applying cutting-edge technologies such as AI, LLMs, and AI agents to real products and business operations. This includes initiatives such as:

- Building multimodal knowledge graphs

- Enabling LLM/agent processing of large and complex documents

- Developing automated evaluation and self-improving accuracy frameworks for AI agents

- Designing agent systems capable of automating a wide range of business processes

#### Responsibilities

- Research, validate, and productionize applied technologies for integrating AI, LLMs, and AI agents into products.

- Conduct R&D for products that leverage LLMs in practical, business-oriented ways, generating real business value from state-of-the-art technologies.

- Elevate the overall technical capabilities of LayerX. By delivering impactful results, Senior R&D Research Engineers expand the range of technologies available for product development across the company.

- Collaborate not only with Ai Workforce but also with other divisions-including Bakuraku and Corporate teams-to identify the most strategically important research areas for LayerX, define key challenges, and drive solutions forward.

#### What Makes This Position Attractive

- Gain hands-on experience solving real-world business and technical challenges through the practical application of AI and LLM technologies.

- Apply AI, LLMs, and AI agents in production environments. Rather than working solely on academic research, you will tackle real production data and real business operations head-on, solving real-world problems through research and engineering.

Required Qualifications / Experience

- Ability to understand Japanese business documents and communicate effectively in both technical and business contexts.

- Ability to read and understand English research papers in AI, LLMs, generative AI, AI agents, and related fields, and to reproduce implementations and conduct practical evaluations.

- Strong, readable coding skills in Python.

- Ability to appropriately leverage AI coding tools (e.g., Claude Code, OpenAI Codex, OpenCode, Cursor, Devin, GitHub Copilot) and conduct code reviews.

- Hands-on experience independently researching, developing, and utilizing generative AI technologies such as LLMs and AI agents, including (but not limited to):

- Development experience with RAG, AI agents, multi-agent systems, and agentic workflows

- Experience using MCP and A2A

- Experience with LLM APIs such as OpenAI API and Gemini API

- Development and utilization of multimodal AI (e.g., VLMs), not limited to LLMs

- Understanding and practical experience with reinforcement learning, search algorithms, and optimization algorithms

- Experience developing LLM-as-a-Judge systems, hallucination mitigation strategies, and LLM benchmarking frameworks

- Experience using LLM observability and management systems such as LangSmith and LangFuse

- Experience building machine learning models using deep learning frameworks such as PyTorch

- Experience across data domains for AI/ML, including data analysis, data science, data engineering, and feature engineering

- Experience building knowledge graphs and graph databases

- Experience collaborating with customers and internal stakeholders to productionize new technologies and create tangible value in products and business.

- Master’s or Ph.D. degree in Computer Science, Informatics, Machine Learning, or a related field, or equivalent practical experience in corporate R&D.

Preferred Qualifications / Skills

- Business-level communication skills in both Japanese and English.

- Experience in software and system development, including:

- Web application development and operations

- Infrastructure development and operations, including cloud environments

- Data platforms and machine learning infrastructure development and operations

- Development of general-purpose libraries

- Experience with patent applications and inventions.

What We’re Looking For

- Alignment with LayerX’s mission: “Enabling digitalization across all economic activities.”

- Resonance with our core values: Virtue (徳), Trustful Team, Bet AI, Fact Base, Be Animal

- Shared understanding of and alignment with [LayerX羅針盤 / the LayerX Compass (Rashinban)](https://speakerdeck.com/layerx/compass_202209)

Salary

Minimum Annual Salary：From ¥10,000,000 per year

※Please note that the above figures are indicative only, and we are open to offering a higher compensation based on your qualifications and experience.

Development Environment

- **Backend**: Python (FastAPI), GraphQL

- **Web Frontend**: TypeScript, React (Next.js)

- **Database**: Azure Cosmos DB for PostgreSQL

- **Middleware**: Azure OpenAI, Azure AI Search

- **Infrastructure**: Azure

- **Provisioning Tools**: Terraform, Azure Bicep

- **Backlog Management**: Notion

- **Source Control**: GitHub

Reference Articles

- [Ai Workforce Engineering Hiring Deck](https://speakerdeck.com/layerx/ai-workforce-engineering-hiring-deck)

- [Ai Workforce Division Careers Page](https://layerx.co.jp/jobs/aillm/)

- [Ai Workforce Division Deck](https://speakerdeck.com/layerx/aillm-division-deck)

- [Ai Workforce](https://getaiworkforce.com/)

- [https://tech.layerx.co.jp/entry/2025/05/01/110000](https://tech.layerx.co.jp/entry/2025/05/01/110000)

Employment Type・Work Style

#### Employment Type

- Fulltime

- Probation Period : 3 months (with no change in salary and benefits)

---

#### Work Style

- Working Hours : Flextime system with no core hours

- Work Location : Tokyo Office or Remote

---

#### Other Benefits

- Commuting expenses fully reimbursed

- No smoking inside the building

----

- [LayerX's Way of Working in 2026](https://note.layerx.co.jp/n/n4ca3a768c9bf)

DE&I Policy

- LayerX strives to be a place that provides opportunities to people from all diverse backgrounds. During the recruitment process, factors such as age, gender, nationality, race, disability, LGBTQ+ status, employment type, or marital status are in no way taken into consideration. We actively seek team members who share our mission and resonate with the business we operate.

---

- [LayerX DE&I Policy](https://layerx.co.jp/about/deandi/)

└LayerX has established a DE&I Policy. For the background behind its formulation, please read the note by our CEO Fukushima ( https://note.com/fukkyy/n/nc0b6195f5fae )

Location / address

- Tokyo Office

- 5th Floor, Ginza Shochiku Square, 1-13-1 Tsukiji, Chuo-ku, Tokyo

Benefits

- Support for visa acquisition and renewal

- Five-day workweek with full weekends off (including public holidays, year-end and New Year holidays)

- Complete social insurance coverage

- Maternity, childcare, and caregiving leave systems

- Financial support for participation in conferences and study sessions

- Welcome leave: 3 days of paid leave granted upon joining, in addition to the paid leave granted after six months

- Life support leave: 4 days of paid leave granted upon joining, separate from the leave granted after six months (*subject to eligibility)

- Refresh leave: 3 days of paid leave granted to employees enrolled as of June 30th

※In addition, we encourage flexible work arrangements tailored to each member’s situation. Please feel free to discuss any concerns during the recruitment process.

Recruitment Process

#### Main Selection Process

- (Casual interview) → Document screening → 1st interview → 2nd interview → Final interview → Offer discussion

---

- After confirming your intention to proceed, you will have approximately 1 to 3 interviews before the final interview.

- During the process, you may meet with multiple team members.

- Interviews are primarily conducted online. Upon request, offline communication such as office tours can be arranged.

- The selection period is typically around 3 to 5 weeks. Depending on the candidate’s situation, we may arrange a condensed schedule, possibly completing the process in as few as 2 days to accommodate your preferences.

- Interviews will be conducted in English. However, scheduling and related communications may be handled in Japanese, so please use translation tools as needed.

- A reference check will be requested in the final phase. We would appreciate it if you could prepare two referees. For more details, please refer to [this note](https://layerx.notion.site/c564813dc0364eb5be20374614c8fc75)

LayerX Information

- [LayerX Hiring Information](https://jobs.layerx.co.jp/)

└Latest and related information about LayerX’s recruitment is compiled here. We kindly ask you to review it before proceeding with the selection process.

- [LayerX CompanyDeck](https://layerx.page.link/company-deck)

└We provide information about LayerX’s business overview, core values and culture, as well as our teams.

- [LayerX Engineer's blog](https://tech.layerx.co.jp/)

└This is the blog of LayerX’s development team, where we share updates on technical initiatives and insights about our development team and work style.

- [LayerX NOW!](https://anchor.fm/layerx)

└This is LayerX’s daily life podcast. All the content is currently available only in Japanese.

- [CEO Fukushima's note](https://note.com/fukkyy/)

└This is a note by LayerX CEO Fukushima, where he shares his passion for LayerX from various perspectives.