LegalOn Technologies, known for developing the AI legal platform “LegalOn Cloud,” launched a new product called “CorporateOn” in January 2025. CorporateOn is an AI-powered assistant that helps employees find answers about internal regulations, manuals, and other corporate policies.

Just like laws help keep society running smoothly, LegalOn has been building AI shaped by  deep legal expertise. We sat down with the team to hear more about their move into the corporate space and how they’re thinking about the future of AI in the workplace.

Yuki Yoshinaga | Product Designer / Product Manager – CorporateOn/ LegalOn Technologies, Inc.

After working as a freelance UX designer and UX researcher across various projects, Yuki ventured into entrepreneurship. He successfully sold a B2C app business and managed a D2C brand before joining LegalOn Technologies in 2025. At LegalOn, he leads design, research, and product management for new business initiatives, including CorporateOn, a solution for corporate departments.

Jun Nomura |  Product Designer – LegalOn Technologies, Inc.

Jun began his career at a design agency, focusing on UI design with an emphasis on prototyping. He later worked as a freelancer and at an e-commerce platform company before joining LegalOn Technologies in March 2024. In his current role, he is responsible for design and user research for LegalOn’s contract management tools.

CorporateOn, a tool that supports greater efficiency in corporate operations

To start off, could you please tell us about LegalOn Technologies and its product, “CorporateOn”?

Jun: LegalOn Technologies is a global AI company with deep expertise in both legal knowledge and advanced AI development. Our main product, “LegalOn Cloud”, is an AI-powered platform that helps streamline contract management. It brings together features like contract review and post-signature support, which were previously offered through our earlier products, LegalForce and “LegalForce Cabinet”.

Yuki: Released in January 2025, CorporateOn is a new product designed to serve a wider range of corporate needs. Its main feature allows users to upload documents such as PDFs to make internal rules, manuals, and policies searchable. Based on those materials, CorporateOn automatically generates answers to employee questions.

Yuki: In corporate departments, it’s common to hear employees ask, “How should we handle this?” or say, “I still don’t get it, even after reading the rules.” Answering these questions can take up a lot of time, leaving less room for more important work. It can also be stressful to deal with the same inquiries over and over again. By having CorporateOn handle these kinds of questions, we aim to help teams use their time more effectively.

At first, we didn’t plan to create a product specifically for corporate departments. Our main focus was actually a product to support decision-making for executives.

I see! What led you to pivot and eventually develop CorporateOn?

Yuki: During research and business discussions about the executive focused product, I had several conversations with executives in corporate departments. From these talks, we uncovered insights that shaped the current concept.

We also realized that LegalOn Technologies already has many team members who are domain experts, lawyers, certified public accountants, tax accountants, and management consultants who have direct experience in corporate areas like finance, HR, IT, and corporate planning. Leveraging this in-house expertise gave us confidence we could build a product that truly understands user needs.

Jun: I’m part of a different development team from the one working on CorporateOn, so I observed the shift from the outside. I was really surprised when the direction changed so drastically. I wondered, “They were so passionate about it, what happened?” But I realized it was the result of learning from the real needs of our customers.

I think it’s a great success story that the frontline team learned from customers and proposed a change in direction to management, resulting in a decision both sides could agree on. It shows that putting your passion into the work while facing customer needs head-on can truly pay off.

Using In-House Knowledge as a Strategic Advantage to Support Employees

During the product development process, how did you gather domain knowledge and insights and integrate them into the product?

Yuki: We began by thoroughly understanding the work of LegalOn Technologies’ domain experts. We’re proud to have exceptionally talented team members, and by learning not only about the advanced roles within corporate departments but also about their future directions, we gained valuable insights.

With permission from our client partners, we also conducted user interviews early in the evaluation phase.

After gathering all this information, we held an event storming workshop to organize and visualize the business workflows in a single diagram. Sharing this diagram across departments ensured every developer fully understood the domain, even down to the smallest details.

*Event Storming is a workshop-style modeling technique proposed by Alberto Brandolini. It focuses on deeply understanding domain events and aims to share, visualize, and comprehend complex business processes and domain knowledge.

What discoveries did you make and apply during the product development?

Yuki: One discovery was that users aren’t simply satisfied with getting answers to questions about company regulations. While FAQ chatbots exist, many users find them unsatisfactory and often don’t use them at all. This happens because corporate teams struggle to find time to update these chatbots while managing their regular duties. Additionally, since these FAQ chatbots are rarely used, it becomes difficult to justify the resources needed to maintain them. As a result, they often end up neglected, creating a negative cycle.

Based on this insight we realized that the product design needed to focus on two key points: minimizing management effort and ensuring employees actually use the product.

To reduce the management burden, we developed a system where users simply upload existing documents. On the usage side, employees do not work just to follow rules, they work to achieve results. That means there is little demand for simply searching rules. Instead, the product needed to offer expert knowledge that helps teams move forward proactively, such as strategic insights or information that supports launching new business initiatives.

What steps did you take to integrate expert knowledge into the product?

Yuki: LegalOn Technologies has gathered extensive knowledge and insights from domain experts, along with deep experience in corporate legal affairs. We consolidated and organized this wide-ranging expertise into a proprietary database called the “Corporate Knowledge Graph”, which we then integrated into the product. CorporateOn is a product made possible through the collaboration of many LegalOn Technologies team members, leveraging our own collective knowledge and insights as valuable assets.

Will LegalOn Technologies’ expertise continue to play a role in future feature additions and improvements?

Yuki: Absolutely. In fact, a feature recently proposed by our own corporate department at LegalOn Technologies was added to the development roadmap. Currently, the corporate team manages inquiries from other departments while also using that information to train the AI. This learning process follows two main paths: sharing knowledge and creating manuals within the team, and developing and updating manuals for use across the entire company.

We saw this as a very positive cultural approach, and have started developing features to incorporate it into CorporateOn.

The key to success isn’t just AI, but the human knowledge and insights powering it

CorporateOn is an AI-powered product. What kind of AI does it use, and what capabilities does it offer?

Yuki: We have built a system called the Corporate RAG System, which uses Retrieval-Augmented Generation AI. It performs searches within a vector database (*), which differs from traditional databases in that it references relationships and attributes of information.

Based on this, a large language model (LLM*) interprets the query to determine where in the multidimensional vector space the desired information is likely located. It then returns the most relevant results from nearby data points.

_{*Vector database: a technology that represents and manages unstructured data in a multidimensional vector space, enabling efficient searches.
*LLM: abbreviation for large language model, capable of performing a wide range of natural language processing tasks.}

Thanks to this approach, CorporateOn can deliver search results that reflect meaning and intent, rather than relying solely on exact keyword matches as traditional search methods do.

Yuki: Currently, many companies store manuals and documents in the cloud and rely on search functions to find what they need. You could say they create their own search environments separate from general search platforms like Google. But honestly, this setup is quite inconvenient. Information piles up, it becomes hard to know where things are, rules get complicated..and in the end, people don’t know what to do or whom to ask. This system is designed to solve those problems.

When using AI, are there any specific principles or company values that guide your approach?

Yuki: Having the technical ability to apply AI where it makes sense is essential for future software development. However, to stay competitive over time, what truly matters is maintaining domain expertise as a corporate asset. This is the knowledge that powers the AI. AI alone offers only a short-term advantage. In the mid to long term, real strength comes from the foundation the AI is built on and the knowledge and resources it draws from.

In this respect, expertise in the corporate field is even broader than in the legal field and often harder to organize. Much of it comes from experience, and AI without this added knowledge cannot generate information about how things were handled in the past. The expert knowledge behind CorporateOn is created by members of LegalOn Technologies. It may be more accurate to say that AI is a tool for delivering the human knowledge and insights people possess.

Maintaining Integrity in Legal Tech by Limiting Content Generation

Are there other initiatives related to AI that you expect to incorporate into future product improvements?

Yuki: One of our key focus areas is patent acquisition. As AI technology becomes increasingly widespread, we are filing multiple patents related to CorporateOn to protect our innovations and make it harder for competitors to copy or imitate them.

We’re also focusing on how we process the data we collect and generate. For instance, data created for CorporateOn may be adapted for use in LegalOn Cloud, and we’ve already started preparing for that potential.

When providing AI-powered services to users, are there any particular points you pay special attention to?

Yuki: We are careful to avoid generating excessive content. Because LLMs generate responses based on training data and publicly available web sources, there’s a risk that the output could unintentionally infringe on copyrights. The more the model generates, the greater that risk becomes.

To prevent this, we limit generation to information that has been cleared from a copyright standpoint, primarily the company’s own knowledge and expertise built into CorporateOn. We also keep output as concise as possible and have implemented several safeguards to ensure everything complies with legal standards.

As a legal AI company, we place a high priority on respecting copyright.

Committed to learning and growing together as one

Finally, could you share your vision for the future?

Yuki:  The year 2025 is being called the Year of the AI Agent. We’re at a turning point, moving from an era where humans give all the instructions and AI simply responds, to one where AI independently decides what to do, evaluates its own outputs, revises them, and repeats this cycle. Ultimately, we aim to build AI agents like this ourselves.

Jun: Even with LegalOn Cloud, our development leader has always wanted to create an AI that quietly helps and supports users. Everyone on the team shares this idea. It’s like the robot helpers you see in sci-fi movies, not just a tool that does tasks automatically or a guide that tells you what to do, but a companion that learns and grows with people. That’s how I see it.

This idea isn’t limited to LegalOn Cloud. We want to apply it across all LegalOn Technologies products and turn them into smarter, more helpful AI agents.

We started by focusing on contracts because they are essential to business success. Just like CorporateOn expanded our work beyond legal, we plan to slowly grow into related areas, helping users reach whatever goals they have.

Yuki: In the corporate world, there is a chronic labor shortage, causing problems to build up like a snowball and making it hard to take on new challenges. That’s why we want to offer AI agents that can handle specific tasks almost as well as humans, helping corporate departments solve these issues.

We also want to support employees so they can move forward with confidence. Since internal policies, manuals, and industry-specific rules are closely connected, our future goal is to have CorporateOn reference all of these to provide even better support.

Special Thanks to:

LegalOn Technologies, Inc.
https://en.legalontech.jp/
https://www.legalontech.com

AI Counsel “CorporateOn”
https://lp.www.legalon-cloud.com/corporateon