Gunn Gu Hwang (ABC Studio ABC Merchant System), Ka Young Lee (LINE Domain Product Development 4), and Dong Hoon Kang (Fintech Engineering)
At LINE Plus, AI Transformation (AX) isn't happening in boardrooms and presentation decks, it’s happening for real, in the everyday code, Slack threads, and morning coffee chats of developers. Enter the AI Evangelists: 20 hand-picked developers officially given 10% of their working hours to explore and implement new AI ideas and share their knowhow. They are pioneers in adapting AI agents into workflows and even their personal lives, and they are here to introduce “AI 101” in LINE Plus.
We sat down with three of our star Evangelists—Gunn Gu Hwang (ABC Studio ABC Merchant System), Ka Young Lee (LINE Domain Product Development 4), and Dong Hoon Kang (Fintech Engineering)—to unlock their secret prompts, cultural hacks, and vision for how LINE is using AI to rewire the way it works.
Meet the AI Evangelists
Think of them as LINE Plus's internal AI influencers. Selected from tech organizations across the company (about 20 members in total), the AX team embodies three important characteristics:
● Practitioners: People who build inventive, productivity-boosting workflows (and whose prompts get stolen by teammates).
● Sharers: Generous souls spreading AI hacks through Slack and live sessions.
● Explorers: Trend-seekers constantly hunting for new tools to test.
The program brings together employees who are already experimenting with AI in their daily work and gives them an official mandate to spread what works. They are practitioners who have changed their own workflows, sharers who help colleagues adopt new approaches, and explorers who keep testing what comes next.
More importantly, their work is not treated as an after-hours hobby. AI Evangelist activities are recognized as part of their jobs, with roughly 5–10% of working time allocated to consulting, sharing use cases, running workshops, researching tools, and helping teams rethink how work gets done.
In other words, LINE Plus is not simply asking employees to “use more AI.” It is creating people inside the organization whose job is to make AI useful.
‘I’m not here to cook meals for you, I’m here to show you the kitchen’
Q. How do you explain your job as an AI Evangelist to your family or non-tech friends?
Dong Hoon: I use a food analogy! I tell them: “I’m not here to cook meals for you, but to show you what’s inside the kitchen.” I can't do other people's work because only they truly know their domain. My role is just to spark that “Aha!” moment where they realize, “Wait, I can make AI do this tedious part for me!”
Ka Young: I’m someone standing at the front line of AX, helping the whole team level up!
Gunn Gu: People face a huge psychological hurdle with AI at first. Getting over that wall takes a friendly nudge. I just see myself as a trusted buddy walking alongside them to cross that threshold together.
Q. If you could only pick ONE important skill an AI Evangelist must have, what is it?
Ka Young: 100% communication! You can be a genius AI wizard, but if you can't translate that magic to your team, nothing changes.
Dong Hoon: Agreed. Communication is the key. AI trends expire every six months anyway. If you just shout “Hey, use this tool!”, people ignore it. But if you dig into their specific pain points and say “Here’s how to fix that headache,” suddenly their eyes light up.
Gunn Gu: Empathy and proactiveness! You need to observe your colleagues like a detective, ask what’s bugging them, and brainstorm fun solutions together.
Dong Hoon: I use a food analogy! I tell them: “I’m not here to cook meals for you, but to show you what’s inside the kitchen.” I can't do other people's work because only they truly know their domain. My role is just to spark that “Aha!” moment where they realize, “Wait, I can make AI do this tedious part for me!”
Ka Young: I’m someone standing at the front line of AX, helping the whole team level up!
Gunn Gu: People face a huge psychological hurdle with AI at first. Getting over that wall takes a friendly nudge. I just see myself as a trusted buddy walking alongside them to cross that threshold together.
Q. If you could only pick ONE important skill an AI Evangelist must have, what is it?
Ka Young: 100% communication! You can be a genius AI wizard, but if you can't translate that magic to your team, nothing changes.
Dong Hoon: Agreed. Communication is the key. AI trends expire every six months anyway. If you just shout “Hey, use this tool!”, people ignore it. But if you dig into their specific pain points and say “Here’s how to fix that headache,” suddenly their eyes light up.
Gunn Gu: Empathy and proactiveness! You need to observe your colleagues like a detective, ask what’s bugging them, and brainstorm fun solutions together.
‘Don't Hesitate to Refute Me’: AI Hacks & Prompts
Q. What is your go-to "magic prompt" or secret trick when working with AI?
Dong Hoon: Oh, I love giving AI a reality check. My favorite prompt: “Do not blindly modify code just because I pointed something out. Think dispassionately. Don't hesitate to say NO to me if it's rational, but don't blindly oppose me either.” If you criticize AI, it instantly apologizes even if you were wrong—you have to tell it to fight back!
Gunn Gu: I use AI as a devil's advocate. My prompt is: “Falsify my instructions based on critical thinking.” If you only ask AI for validation, it just echoes your own bias. Using AI to smash your assumptions saves massive company time before you ship code.
Ka Young: Mine is simple but essential: “If anything is ambiguous in my request, STOP and ask me first before assuming!”
Dong Hoon: Oh, I love giving AI a reality check. My favorite prompt: “Do not blindly modify code just because I pointed something out. Think dispassionately. Don't hesitate to say NO to me if it's rational, but don't blindly oppose me either.” If you criticize AI, it instantly apologizes even if you were wrong—you have to tell it to fight back!
Gunn Gu: I use AI as a devil's advocate. My prompt is: “Falsify my instructions based on critical thinking.” If you only ask AI for validation, it just echoes your own bias. Using AI to smash your assumptions saves massive company time before you ship code.
Ka Young: Mine is simple but essential: “If anything is ambiguous in my request, STOP and ask me first before assuming!”
Q. What has been the biggest game changer in your personal daily workflow since AI took over?
Dong Hoon: I stopped saying “I'll look into it later” for things that I do not know. In the past, heavy research tasks sat on my to-do list forever. Now, I set AI to run research in the background while I do other tasks. Also, building quick internal tools has become so cheap that I just spin them up on the fly whenever I get annoyed by a manual task.
Gunn Gu: Slack thread overload? Solved. I have AI read, summarize, and structure long threads into neat summaries. Documentation happens almost automatically in real-time.
Ka Young: Overcoming “blank page syndrome”. Even daunting tasks feel easy because I just draft a messy idea and refine it with AI right away.
Before AI: “I’ll Look Into It Later.”
After AI: “Let’s Just Check Now.”
[Inside the Engine] How the Magic Spreads
From 1-on-1 coffee chats to all-hands hackathons—The Evangelist mission isn't just about personal efficiency, it’s a full-spectrum community operation:
● Team Level: Quarterly 1-on-1 AI consulting sessions tailored to each team's tech stack, hands-on DLC workshops, and mini-hackathons.
● Company-Wide Level: Lively Slack chats, bi-weekly Evangelist meetups (with real-time policy feedback pushed straight to DevRel), and the legendary “After Talk” networking sessions where developers chat about AI over drinks.
● Workshop for Cross-Functional Collaboration: 18 AI Evangelists designed and facilitated the AI DD Workshop, bringing together 62 participants across Project Management, Design, QA, and Engineering from 13 product teams. The two-part program—spanning AI Setting Day (hands-on AI readability, skills, and context missions) and AI Inception Day (collaborative drafting of specs, stories, and prototypes)—achieved a 91.2% positive rating, with 53.2% of participants progressing to advanced AI adoption levels as cross-functional handoffs shifted from solo prompt usage to a shared, team-wide product delivery process. Pre-event surveys showed that 20–30% of participants said they had never automated a task, yet 80–90% ended up building a working script on the spot.
From 1-on-1 coffee chats to all-hands hackathons—The Evangelist mission isn't just about personal efficiency, it’s a full-spectrum community operation:
● Team Level: Quarterly 1-on-1 AI consulting sessions tailored to each team's tech stack, hands-on DLC workshops, and mini-hackathons.
● Company-Wide Level: Lively Slack chats, bi-weekly Evangelist meetups (with real-time policy feedback pushed straight to DevRel), and the legendary “After Talk” networking sessions where developers chat about AI over drinks.
● Workshop for Cross-Functional Collaboration: 18 AI Evangelists designed and facilitated the AI DD Workshop, bringing together 62 participants across Project Management, Design, QA, and Engineering from 13 product teams. The two-part program—spanning AI Setting Day (hands-on AI readability, skills, and context missions) and AI Inception Day (collaborative drafting of specs, stories, and prototypes)—achieved a 91.2% positive rating, with 53.2% of participants progressing to advanced AI adoption levels as cross-functional handoffs shifted from solo prompt usage to a shared, team-wide product delivery process. Pre-event surveys showed that 20–30% of participants said they had never automated a task, yet 80–90% ended up building a working script on the spot.
AI DD Workshop
The Shift: When AX Gets Real
Q. When was the exact moment you thought, "Wow, our office culture is REALLY changing"?
Dong Hoon: During our hands-on workshop. Most of the 53 participants showed up feeling completely lost about automation. But after seeing a few simple examples, something clicked. By the end of two hours, 47 out of 53 people had built their own automated workflow! That was pure magic.
Gunn Gu: When non-engineering project planners started dropping by to ask how to use AI tools and began building stuff on their own. That’s when I knew the walls were coming down.
Ka Young: When team managers and leaders started actively coming to us asking, “How can we bake AI deeper into our team sprint?”
Q. What has been the most popular AI automations built by employees so far?
Dong Hoon: You won't believe it—the top-voted script was: “Automatically booking a meeting room on the floor I checked in for the day.” Forget complex system architecture; people love solving those tiny, hilarious everyday annoyances. Relatable practicality wins every single time!
Gunn Gu: “For me, I don’t read any slack messages anymore. Instead, I ask AI to read them and summarize them for me. This really saves my time by speeding up the documentation and also making sure I don’t miss out on any messages.
Dong Hoon: During our hands-on workshop. Most of the 53 participants showed up feeling completely lost about automation. But after seeing a few simple examples, something clicked. By the end of two hours, 47 out of 53 people had built their own automated workflow! That was pure magic.
Gunn Gu: When non-engineering project planners started dropping by to ask how to use AI tools and began building stuff on their own. That’s when I knew the walls were coming down.
Ka Young: When team managers and leaders started actively coming to us asking, “How can we bake AI deeper into our team sprint?”
Q. What has been the most popular AI automations built by employees so far?
Dong Hoon: You won't believe it—the top-voted script was: “Automatically booking a meeting room on the floor I checked in for the day.” Forget complex system architecture; people love solving those tiny, hilarious everyday annoyances. Relatable practicality wins every single time!
Gunn Gu: “For me, I don’t read any slack messages anymore. Instead, I ask AI to read them and summarize them for me. This really saves my time by speeding up the documentation and also making sure I don’t miss out on any messages.
The Golden Rules of AX
Q. Final advice: What is the real secret to spreading AX across a large tech company?
Dong Hoon: Two things: Lower the bar and make it visible. People think automation needs to be this grand, permanent architecture. Nope—a dirty script you use once and throw away is still automation. And once you build it, share it so the person sitting next to you can copy-paste it.
Ka Young: Never forget that humans stay in the pilot seat. AI is your super-talented co-pilot, but accountability always stays with you.
Q. Fill in the blank: "To me, AX is __________."
Gunn Gu: AX is a tool for falsification—the fastest way to break my own confirmation bias.
Ka Young: AX is a shiny gift that lets me accomplish things I couldn't even dream of before.
Dong Hoon: AX is JUST DO IT. For me, AX means trying before judging. Stop overthinking costs—just build it first, test it in minutes, and evaluate later. That flipped mindset changes everything.
Dong Hoon: Two things: Lower the bar and make it visible. People think automation needs to be this grand, permanent architecture. Nope—a dirty script you use once and throw away is still automation. And once you build it, share it so the person sitting next to you can copy-paste it.
Ka Young: Never forget that humans stay in the pilot seat. AI is your super-talented co-pilot, but accountability always stays with you.
Q. Fill in the blank: "To me, AX is __________."
Gunn Gu: AX is a tool for falsification—the fastest way to break my own confirmation bias.
Ka Young: AX is a shiny gift that lets me accomplish things I couldn't even dream of before.
Dong Hoon: AX is JUST DO IT. For me, AX means trying before judging. Stop overthinking costs—just build it first, test it in minutes, and evaluate later. That flipped mindset changes everything.
The experiences of these AI Evangelists proves one thing clearly: true transformation doesn't come from top-down speeches, but from the bottom-up passion of developers having fun with technology. They are not trying to make everyone an AI expert. They are trying to make AI ordinary.
Ordinary enough that a developer no longer postpones investigating something because it will take too long. Ordinary enough that someone in a workshop realizes a repetitive annoyance can be automated and builds a solution in a couple of hours. Ordinary enough that an individual experiment becomes a team skill. And ordinary enough that using AI is no longer an extra activity on top of work—but simply part of how work gets done.
That is where LINE Plus’s approach to AX becomes less about technology and more about culture. By putting that culture first, we ensure that all our technology and products are uniquely LINE, always focused on putting users first. And thanks for our AI Evangelists, this next chapter in LINE’s journey will redefine how our tech teams build, collaborate, and grow.
Ordinary enough that a developer no longer postpones investigating something because it will take too long. Ordinary enough that someone in a workshop realizes a repetitive annoyance can be automated and builds a solution in a couple of hours. Ordinary enough that an individual experiment becomes a team skill. And ordinary enough that using AI is no longer an extra activity on top of work—but simply part of how work gets done.
That is where LINE Plus’s approach to AX becomes less about technology and more about culture. By putting that culture first, we ensure that all our technology and products are uniquely LINE, always focused on putting users first. And thanks for our AI Evangelists, this next chapter in LINE’s journey will redefine how our tech teams build, collaborate, and grow.