I combine enterprise technical coordination, training, workflow analysis, documentation, and practical AI prototyping to help organizations introduce AI within clear human, data, and governance boundaries.

My current work focuses on practical AI adoption, workflow transformation, responsible AI, AI literacy, pilot evaluation, and human-centred agent design.

I combine enterprise technical coordination, training, workflow analysis, documentation, and practical AI prototyping to help organizations introduce AI within clear human, data, and governance boundaries.

My current work focuses on practical AI adoption, workflow transformation, responsible AI, AI literacy, pilot evaluation, and human-centred agent design.

I bring experience in enterprise technical support, international knowledge transfer, technical training, research computing, documentation, and cross-functional coordination.

At OpenText, I supported collaboration between the Tokyo office and Waterloo headquarters across Technical Support, R&D, Product Management, Professional Services, Sales, partners, and regional stakeholders. I also developed and delivered technical learning materials in Japanese and English.

In university and research environments, I supported scientific data workflows, created procedural documentation, and explained technical concepts to researchers with different levels of programming experience.

Comparative research and practical guidance for institutional AI adoption, governance, literacy, pilot design, human oversight, and the responsible use of protected institutional information.

This repository contains reports and planning resources primarily for higher-education leaders, AI advisory and enablement teams, and privacy, security, legal, data-governance, procurement, accessibility, records-management, and risk professionals.

Topics include:

• launching an institutional AI advisory, literacy, and pilot service;
• institutional control architecture and governance roles;
• service tiers based on information sensitivity, persistence, integration, and impact;
• pilot lifecycle and approval gates;
• repository-connected AI and permission risks;
• principles for responsible AI adoption and governance;
• pilot evaluation before scaling;
• human accountability, traceability, and periodic review.

A testable Agent Skill for evaluating proposed AI-assisted workplace workflows.

The framework reviews:

• workflow evidence and organizational relevance;
• appropriate tool selection;
• pilot readiness;
• human-review requirements;
• failure cases and exceptions;
• employee workload, trust, and recognition;
• governance and accountability;
• organizational knowledge accumulation;
• stop, revise, escalate, or scale decisions.

The project includes:

• a SKILL.md;
• supporting reference files;
• realistic organizational examples;
• evaluation cases with expected outcomes;
• explicit limits and human-approval requirements.

A beginner-friendly, offline, static web tutorial that shows how to create a personalized, user-controlled AI context system using Microsoft Copilot.

No coding is required.

The tutorial includes:

• a seven-stage guided workshop with a safety checkpoint;
• copyable prompts;
• a Microsoft Copilot guide based on verified documentation;
• a fully offline design with no server, API calls, or CDNs;
• practical guidance for organizing personal context while retaining user control.

• AI adoption and transformation
• Responsible AI and governance
• AI literacy and enablement
• Workflow analysis
• Human-in-the-loop systems
• Knowledge management
• AI evaluation and pilot design
• Organizational learning
• Data and access governance

My work is guided by several recurring principles:

1. Start with real workflow evidence rather than a preferred tool.
2. Decompose work before deciding where AI may help.
3. Use narrow, reversible pilots with measurable baselines.
4. Define human review as part of the workflow.
5. Document failures and exceptions as reusable knowledge.
6. Separate advice, approval, and automated action.
7. Use the minimum information required for the task.
8. Evaluate quality, workload, trust, accessibility, and risk alongside efficiency.
9. Revalidate systems as models, data, permissions, and organizational conditions change.

I am currently developing practical resources for:

• institutional AI literacy and advisory services;
• protected experimentation and pilot support;
• tool, data, and interaction decision frameworks;
• role-based AI training;
• human-review design;
• repository-connected AI controls;
• evaluation and scaling criteria for AI-assisted workflows.

• LinkedIn: Nobuki Matsui