It looks like smooth sailing from the outside, but underneath, we’re navigating a storm.

Esther once believed that "AI companies should be avoided." This was because there were too many places that were merely wrapped in plausible packaging without any real technological depth.

As a result, her standards became much higher: an early-stage setup, zero to one, and problems that no one else had solved. Twelve Labs was the choice that remained after passing through those strict criteria.

Today, as Director of Engineering, she leads teams across the US and Korea, handling everything from recruitment to technical decision-making.

Q. You often introduce Twelve Labs as a "video AI company." From an engineering perspective, what do you see as the core of this work?

I actually think the AI aspect is a much bigger factor than the video itself. No matter how complex a typical application service is, it remains somewhat deterministic. The same input should yield a similar result, and you can design a system on that assumption. 

However, everything changes when it is AI-based. Even if you ask the exact same question, the answer can differ between today and tomorrow. Building products and running engineering based on that unpredictability is a unique class of problem that is rarely encountered in traditional IT domains.

At Twelve Labs, we add the complexity of video data to the inherent uncertainty of AI, giving us an identity that is difficult to find anywhere else.

Q. That makes me wonder why you tackle such a difficult challenge. Was there a specific moment when you felt this technology was truly meaningful?

Behind the shorts, reels, and streaming recommendations we watch daily, there is ultimately a process of understanding, categorizing, and connecting video. These are tasks that are far too vast, slow, and prone to human error if done manually.

What we do is enable people to utilize multimodal data (video, image, audio) with peak efficiency. This means automatically extracting metadata, classifying content more accurately, and helping recommendation systems work better. It serves as the foundation for how people discover and consume content.

While it operates behind the scenes, it directly shapes the way the media world moves. I believe that is the true meaning of this technology.

Q. What kind of person do you think fits well with Twelve Labs?

In a startup, work doesn't get divided cleanly into neat boxes of 'my job' versus 'your job.' Because of this, people who prefer to solve problems as soon as they see them tend to fit better than those who enforce rigid role boundaries.

In larger organizations, even tiny changes must go through multiple processes, and you often hear, "That's not your role." Here, the attitude is much closer to "If you see a problem, let's fix it quickly." If you are someone who thrives in an environment where you can make an immediate, tangible impact, you will fit right in.

Q. You mentioned that you originally wanted to avoid AI companies. Why did you choose this place?

People around me tried to talk me out of it. There really were many companies back then that had nothing but hype without any real technology. The phrase "AI is a scam" almost sounded like a fair point.

But I wanted to join an early-stage startup, and specifically, I wanted to experience zero to one. And there was one more condition: it had to be a problem that no one else had solved yet. E-commerce or similar fields would have been fine, but working on problems everyone was already addressing sounded boring.

Entering my 30s, I realized that the energy of your 20s is different from your 30s, and there are challenges you can only take on at specific stages of life. I decided that I needed to take the kind of leap now that I might not be able to in my 40s. With that standard in mind, Twelve Labs was the perfect fit.

Q. When you work, there are moments when things go off track or attempts don't pan out. How do you approach those moments?

I don't think failure is such a big deal. I believe that mistakes can always happen as long as they can be resolved. People make mistakes, AI makes mistakes, and machines make mistakes.

What matters is whether we learn from it. While we have structured processes like post-mortem meetings, I believe a culture of self-reflection is far more critical. Asking yourself what you could have done better, and giving feedback to your manager if there are areas where the company needs to improve. When that reflection accumulates, it builds a solid system and prevents the same issues from recurring.

Q. What are the people at Twelve Labs like?

In a word, we don't have Doberman personalities; people here are more like Golden Retrievers. When a problem arises, instead of pointing fingers, everyone runs over together saying, "Oh no! How do we fix this?" It's a rare dynamic to see in an engineering organization, but over time, it has truly solidified as our culture.

Startups have plenty of challenging moments. The workload can get heavy, or things might not go smoothly. At the end of the day, what keeps you going is the people beside you. If you compare it to a tree, we are a team with deep, strong roots. We aren't floating around without support; we might sway, but we won't easily fall.

Q. Is there something you feel you wouldn't have learned if you weren't here?

I would have never experienced hiring and managing engineers in the US, especially leading a team working across such different time zones for the first time.

Additionally, because Twelve Labs is a globally recognized company in the AI space, it opens up unique touchpoints. Working with tech leadership from other companies and engaging in diverse partnerships are opportunities that naturally come our way—things that would have been hard to access if I had stayed solely within Korea. That proved to be an unexpected accelerator for my professional growth.

Q. Management is never an easy job. What are the moments that give you energy?

It's seeing the team grow. I love watching existing members step up, and it's exciting to see new hires quickly integrate and become "Twelvies." Interestingly, almost no one at Twelve Labs feels disappointed after joining. People hear good things from the outside, but once they arrive, they often say it's even better than they expected. Seeing that reaction makes me feel that building this culture together was incredibly meaningful.

Q. Regarding AI itself, what is the most interesting aspect for you personally?

AI always brings surprises. Sometimes it works perfectly as intended, but other times it develops in unexpected ways. It's like a mystery box.

You might think it will rank first in class because you trained it heavily, but it ends up ranking tenth, or you might train it less and one day it suddenly leaps forward. It feels very human. That's why I think understanding the limitations of AI is actually more important to utilize it well. People often start with "Let's leave it all to AI" and eventually realize this truth. Finding and setting those boundaries seems to be the most critical question in our field right now.

Q. Lastly, is there anything you'd like to say to engineers considering joining?

From the outside, it might look like a calm voyage. But inside, it is a storm. There are many difficult problems, and it's not always smooth sailing. It's not just smiles every day; sometimes relations get tense, and questions like "Why are we doing this?" arise.

But the important thing is that this ship isn't just drifting. The direction we are heading is clear, and everyone is moving toward the same goal. In the process of weathering that storm together, our ship is growing, our crew is expanding, and our teamwork is falling into place. We are definitely not just wandering in the open sea.

If you want to make an immediate contribution, if you want to solve problems that no one else has cracked, and if you want to grow with high density alongside great people, this is the right place for you. 

I look forward to embarking on this voyage with you.

Esther is the Director of Engineering at Twelve Labs. Twelve Labs is currently looking for talented engineers to join our journey → twelvelabs.io/ko/careers