Here at Mercari, most members have already started using AI in their daily work as we move toward becoming AI-Native.

But what exactly do we mean by becoming “AI-Native”? What kind of mindset or type of employee does “AI-Native” refer to?

To shed light on how Mercari’s AI-Native professionals are making the most of AI, we have started a new interview series called “Yurino’s AI Question Corner,” featuring Yurino Horiuchi (@Yurino)—a Mercari intern who currently works on the AI Task Force and will be joining us as a permanent member soon!

For our third installment in the series, @Yurino had a chat with @yukita from BizDev.

After gaining experience in data science at a manufacturing company, @yukita joined Mercari Group as a data analyst and is now in charge of business development related to logistics.

Drawing on a whole toolkit of skills from his days as a data analyst, he puts AI to work in all sorts of ways—from analyzing contracts to building mockups. So, what kind of mindset does he think we really need in the age of AI? Let’s dive into the unique BizDev techniques that he uses to develop AI right alongside his projects.

---

---

Developing AI for each project for context-driven brainstorming

@Yurino: @yukita, how do you usually use AI?

@yukita: I mainly use Cursor in my work. For each project, I set up a repository where I keep research materials, notes from bouncing ideas off AI, data for analysis, and rough drafts of proposals as we move things forward. BizDev work covers everything from market research and data analysis to business negotiations, designing on-site operations, drafting contracts, planning how to handle inquiries, and even creating landing pages. With projects often running for months at a time, building up context along the way becomes absolutely crucial. Cursor helps me move multiple projects forward at the same time.

@Yurino: So Cursor serves as a partner running the project together with you. It sounds like the way information keeps building up makes Cursor perfect for BizDev—you can bounce ideas around from research all the way to brainstorming!

Tapping into a data analyst toolkit

@Yurino: Can you share more details about how you use Cursor in BizDev?

@yukita: Take contracts, for example. If the company signed a basic contract 10 years ago, and since then it’s had all sorts of memoranda added to it, even humans are going to have a hard time interpreting it. But if you hand all your past contracts to Cursor and give it instructions, like “Convert these Word files into an AI-friendly format,” “Sort out what’s changed in this memorandum,” or “Draft a new memorandum on this topic while keeping the previous writing style in mind,” it’ll handle everything for you—including writing and running Python scripts if needed. We also use it for data analysis in the same way—Cursor will take your data, transform it into an easy-to-analyze format, and then create insightful reports.

@Yurino: Sounds like your experience as a former data analyst is really shining through.

@yukita: I think it really helps. From my experience as a data analyst, I often look at a task and get the feeling that I could probably handle it using Python. But honestly, whipping up some Python code just for a one-off task didn’t seem worth the trouble. Now, with Cursor, all I have to do is tell it what I want, and it takes care of the rest—so I can easily do things I used to just give up on. What’s important is to have a toolkit of methods at your disposal—knowing, for example, that certain approaches can be used in specific situations. If you know what approaches are out there, you can just ask AI to do it for you. Just keep in mind, though, that you’ll need to be able to judge the quality of AI’s output.

Letting AI handle the first 80% to help the pros really show what they can do

@Yurino: What are some particularly impactful ways your team has been using AI?

@yukita: That would be quickly creating mockups of new service ideas with Figma Make, an AI feature within our design tool. Just tell it what kind of app you want to make, and it’ll whip up a polished design mockup—including all the screen transitions—in no time. Having a concrete design makes it easier for the team to get on the same page, and it also helps everyone spot what the key issues are. When interviewing a user and asking, say, what they think of a certain service idea, it’s always better to show them something as close to the real thing as possible. That way, you’ll have a much better interview.

For example, here’s what a design made with Figma Make might look like. (Note: This is a sample for the article, not an actual service being considered.)

@Yurino: Now that AI can do so much, some people are wondering if AI might end up taking over certain jobs. What do you think, @yukita?

@yukita: Rather than having our jobs taken over, I think our roles will simply change. Sure, AI can probably get you about 80% of the way there. But you know, being a pro is all about taking that 80% and polishing it up to a perfect 100%. When it comes to obsessing over every detail to create something truly great and making the final decision to move ahead, that’s all on us humans. So, personally, I always think: never underestimate the pros! (laughs) For example, whether I’m creating a landing page or requesting a legal review, I always make a draft with AI first and then consult a professional. Up until now, I used to ask the pros to handle the entire process, starting from scratch. But now, I let AI take care of the first 80%, so the pros can really focus on perfecting it to 100%. This way, they can demonstrate their expertise where it’s really needed, and we can move the project forward more quickly. It’s a good deal for both of us, right?

@Yurino: Is there anything you’d like to try with AI in the future?

@yukita: In the short term, I want to make sure that context-aware environments like Cursor repositories aren’t just being developed for me alone—they should be something I can share with other team and project members. On a mid- to long-term horizon, I’d like to take on the challenge of building AI agents that can boost BizDev productivity hundredfold. And rather than having the agents wait for instructions from humans, I want to create a system where AI takes the initiative and sets things in motion on its own. Ideally, I’d love to have an AI that really gets the project’s context in real time and can chime in with suggestions like, “To address the points that came up in this meeting, we should do this research and check in with that person. Should I go ahead and run deep research and reach out on Slack for you?” 

@Yurino: I see, so having AI act as a partner in the thought process, too! Finally, could you share a message to our readers about making the most of AI?

@yukita: First up: just dive in and give it a try! At Mercari, you can use all sorts of tools like ChatGPT, Gemini, n8n, Cursor, and Claude Code. @Yuka and the rest of the Governance Team have set things up so it’s easy and safe to access these tools. Chatting with friends from other companies has made me realize there aren’t many places like Mercari. That’s why I hope everyone enjoys making the most of this unique opportunity.

Secondly: If you’re not sure what you want to do with AI yet, try expanding your horizons first. By reading books outside your own field and coming across ideas for new usages, you might just stumble upon something you want to try doing with AI. Not only does AI make your day-to-day work more efficient, but it also opens up new possibilities—things you couldn’t do before. That’s why I think you’ll actually find it fun!

@Yurino: In other words, when you expand your own toolkit, the rewarding experience of being able to do things you couldn’t before awaits! I think that feeling is the real value of using AI. I’d love to follow your example, too. Thanks so much for sharing your wonderful insights with us today!