“We needed something sharp for one specific job.” — Marcos, Head of of Pre-Production and Localization at InfOhana

There’s a moment in every growing company when the tools stop keeping up with the work.

For InfOhana, that moment came during a localization workflow that should have been straightforward, and wasn’t. Not because the team wasn’t skilled, but because the tools they were using weren’t built for what they were actually doing.

That realization led to something most companies don’t attempt: building their own.

Marcos, who leads Pre-Production and Localization at InfOhana, sat down with us to talk about why the team made that call, what they built, and what it means for the creators and brands they work with.

The Tools Were Good. Just Not for This. 

The localization industry has no shortage of tools. CAT tools, translation management systems, AI-assisted platforms; the market is full of options.

But Marcos was direct about what he found when the team explored them.

“A lot of them are great products. But most of them were built for legal translation, websites, or marketing content, not specifically for YouTube localization workflows. They have features we’ll never use… and at the same time, they’re missing things that really matter to us.”

YouTube dubbing operates by its own rules. Timing has to match a creator’s delivery. On-screen text needs to be accounted for. The voice and personality of each creator (the thing that makes their audience stay) has to carry through into every language.

Generic tools weren’t built around those requirements. And that gap had real consequences for the team.

When Setup Takes Longer Than the Work, Something Has to Change 

Before the internal tool existed, InfOhana’s workflow looked like what most localization teams use: a combination of spreadsheets, video players, and separate documents, switching between them constantly. 

“A translator had to constantly move between different platforms. That kind of setup meant a lot of copy-pasting and context switching, and that’s where mistakes usually happen: inconsistent style and terminology, missed terms, or small details that can affect the final dub.” 

The work always got done. But the setup was costing the team something harder to measure than errors; it was costing them time and energy that should have been going into the translation itself.

“The moment it clicked was when I realized we were spending more time on setup than actually translating. That’s when we said: okay, we need to find or build something that fits the way we work.”

Inside the Workflow: From Script to Final Export 

The workflow starts the moment a video script arrives.

The project manager selects the target languages, and the tool generates an AI-assisted first draft for all languages simultaneously: pulling from the client’s glossary, style guide, and previous translations to ensure consistency from the start.

Translators then work in a clean editor with the video running alongside the script. In real time, the tool checks timing, terminology, and consistency, and flags issues before they become problems in post-production.

Speaker identification and character assignments are handled inside the same environment, which Marcos describes as one of the most underestimated challenges in YouTube dubbing.

“If a line is tricky, the translator can request contextual help from the AI. Once the work is complete, the tool exports everything in the exact format required by the audio team.”

No reformatting. No manual handoffs. No lost context between steps.

Built With the Team, Not Just for the Team

What makes this tool different from a purchased solution isn’t just the features, it’s the process behind them.

“We built it with the team, not just for the team, and that’s an important difference.”

Translators and project managers provide feedback daily. Updates ship almost every week. Some features the team expected to work didn’t, and those were redesigned around how people actually work, not how the tool assumed they would.

“Because we own the tool, we can keep evolving it based on what the team needs.”

That kind of responsiveness isn’t possible with third-party software. A feature request doesn’t go into a product backlog managed by a company that has thousands of other customers. It gets addressed the same week.

Faster Turnaround. Fewer Errors. Full Control. 

The results have been measurable and immediate.

Speed is the most visible change. With everything consolidated into one environment, translators spend more time translating and less time managing the workflow around the translation.

“Before, translators had to jump between different tools. Now everything is in one place, which makes the process much faster and more organized. And in YouTube localization, speed matters a lot, because creators can publish globally faster.”

Consistency has also improved significantly. The tool catches terminology issues and style inconsistencies earlier in the process, which reduces proofing time and improves the quality of what reaches the audio team.

Security is the third pillar, and for Marcos, a non-negotiable one.

“In the YouTube industry, content and intellectual property are extremely valuable. Keeping everything inside our own infrastructure gives us more control and protection.”

What Comes Next

The tool is working. But Marcos is clear that working isn’t the same as finished.

“The biggest limitation is that this isn’t a magic ‘press a button and everything is done’ tool. AI gives us a strong first draft, but human linguists are still the ones who make the translation really good ….and honestly, we want to keep it that way.”

The 2.0 vision focuses on intelligence and personalization; a tool that learns from the work the team does every day, and that increasingly understands each creator’s individual style and workflow.

“Long term, the vision is for more of InfOhana’s processes to run on technology we build ourselves.”

What This Means for Creators and Brands

For the creators and brands InfOhana works with, this infrastructure has a direct impact on what they experience.

Faster turnaround. Stronger consistency across languages. A workflow built specifically around how YouTube content performs and scales internationally, not adapted from tools designed for something else entirely.

“We’re not just offering dubbing services, we’re building the infrastructure behind multilingual YouTube growth. It’s not just about dubbing videos. It’s about building a scalable system for global audience expansion.”

Most localization providers work with the tools available to them.

InfOhana built its own.

Ready to scale your content globally with a team that’s built its workflow around your growth? Reach out to InfOhana and let’s talk about what that looks like for your channel.