Money Forward is expanding AI agents across its products and internal workflows. Reliable agents need relevant context, controlled tool access, and repeatable evaluation.
Within the CTO Office, the Context & Harness Committee develops and validates shared standards for context engineering—the information agents receive—and harness engineering—the execution controls, tool integrations, and evaluations supporting safe, reliable operation.
We are seeking a hands-on AI Tech Lead to lead assigned workstreams, build and test solutions with pilot teams, and turn proven approaches into reusable components and standards for wider adoption.
Expected Mindset
We value rapid hypothesis testing, continuous learning, and technical decisions grounded in product value. You should share findings openly, collaborate constructively, and balance experimentation with reliability, security, compliance, and cost.
Who We’re Looking For
- A hands-on technical leader who enjoys turning ambiguous problems into working, testable solutions.
- Someone who can work across teams without relying solely on formal authority.
- Someone who can distinguish a successful demonstration from a reliable, reusable engineering solution.
- Someone who documents tradeoffs and limitations clearly and helps other teams adopt what works.
Responsibilities
You will combine hands-on engineering with cross-team technical influence, owning the implementation, evaluation, and documentation of your assigned workstream.
Working with committee leadership, pilot teams, and platform and security specialists, you will:
- Lead technical design and delivery for assigned context-engineering or harness-engineering workstreams, clarifying requirements and architecture tradeoffs.
- Review context formats, ontology models, harness designs, and integration approaches.
- Support pilot teams across Money Forward Business Company divisions with context conversion, test fixtures, and evaluation.
- Define how GitHub, Notion, Slack, and other sources supply agent context, including routing and token-budget allocation.
- Build a focused ontology proof of concept using YAML-based schemas.
- Validate agent-platform integrations, including access controls and smoke tests, and demonstrate repeatable evaluation and quality gates.
- Turn pilot evidence into reusable components, playbooks, and recommendations for company-wide standards and adoption.
Requirements
Candidates should demonstrate depth in either context engineering or harness engineering, with working knowledge of the other. The areas below describe the skills relevant to each specialization; candidates are not expected to have equal depth across both.
- Context engineering skills
- RAG and retrieval pipelines: Experience designing advanced search and retrieval systems that provide models with accurate, relevant, and timely information.
- Context-window and token management: Experience compressing conversation history, prioritizing useful context, and filtering noise to optimize model attention and cost.
- Memory architecture: Experience building short-term and persistent memory systems that allow agents to retain state across tasks and interactions.
- Instruction curation: Experience structuring plain-English operating rules, Markdown files, domain guides, and other instructions for dynamic injection into agent tasks.
- Context formats and knowledge modeling: Experience designing context layers and reusable knowledge assets using Markdown, YAML, schemas, and related formats.
- Harness engineering skills
- Tool and MCP integration: Experience connecting models to external APIs, execution environments, tools, and Model Context Protocol (MCP) servers.
- Sandboxing and permissions: Experience defining safety boundaries, access controls, approval flows, and secure execution environments for agent actions.
- Validation and guardrails: Experience implementing linters, automated tests, evaluations, and verification layers that detect hallucinations, policy violations, or broken code.
- Orchestration and control loops: Experience designing multi-step workflows, retry logic, state transitions, and error-correction loops for long-horizon tasks.
- Evaluation harnesses: Experience building fixtures, benchmark tasks, quality gates, and CI-integrated evaluations using DeepEval, Promptfoo, or comparable frameworks.
- Language abilities
- English: Business Level (Fluent, capable of handling communication with clients in English)
- Japanese: N3 Level (Able to communicate internally using translation tools for text and conversation)
Nice to haves
While not specifically required, tell us if you have any of the following.
- Supporting technical and leadership skills
- LLM and agent fundamentals: Strong understanding of agent workflows, RAG, evaluation methods, and quality gates for AI-generated output.
- Software engineering: Strong Git/GitHub and CI practices, experience with repo-native tooling, and the ability to prototype quickly.
- Technical leadership: Ability to make pragmatic architecture decisions, lead POCs, and turn evidence into reusable standards.
- Communication: Ability to write clear technical proposals and provide constructive cross-team review and feedback.
- Experience with ontology or knowledge modeling using YAML-based schemas.
- Familiarity with internal or external agent platforms.
- Experience with hybrid search, indexing pipelines, or context freshness mechanisms.
- Experience in AI development and/or experience in using AI tools to improve development processes. -Money Forward is at a major turning point, shifting “from Cloud to AI.” We are currently driving “AX (AI Transformation)”—the next step beyond DX—with the goal of providing “Digital Workers,” where AI agents autonomously execute tasks. As we enter a phase of evolving into Japan’s No. 1 back-office AI company by integrating AI agents into all of our products in the future, we are looking for individuals who can contribute to AI-driven development and value creation.
Compensation
¥7,908,000 ~ ¥15,000,000 annually.
