In partnership with

Sometimes, the startups we cover have big ideas to help the whole of humanity, while others prefer to drill deep into a specific pain point for a market that most people wouldn’t even consider.

Today’s startup falls firmly in the latter camp. Scroll down to enter the world of voxANN. And sorry for the Police song pun. I really couldn’t resist.

But first:

• We’re partnering with the upcoming Connected North event, and here’s an opportunity for startups:

• Join the Connected North Startup Village in Manchester on April 23-24. Engage with the UK’s top companies, investors, and media. Boost your brand, secure funding, and expand your network across 2,500+ connectivity professionals. ✨ Apply now ✨

– Martin

Free Notion and Unlimited AI

Thousands of startups use Notion as a connected workspace to create and share docs, take notes, manage projects, and organize knowledge—all in one place. We’re offering 3 months of new Plus plans + unlimited AI (worth up to $3,000)! To redeem the Notion for Startups offer:

1. Submit an application using our custom link and select Beehiiv on the partner list.

2. Include our partner key, STARTUP4110P67801.

Learn More

voxANN wants to put on the red light for smoother multi-language audio production

Media production is a mass of niche specialisms. Some of them, like a video editor or a film director, you’ll be familiar with. But have you ever considered the needs of the hard-working people who take content and adapt it for different markets around the world?

Localisation is a big business, whether it’s focused on dubbing a popular children’s cartoon into almost every language on Earth, or making sure the pre-recorded in-flight announcements on a plane are accurate in all the languages an airline serves.

voxANN is a Manchester-based startup looking to serve this global market with a product that makes language service providers’ work cheaper, faster, and easier.

As co-founder Dan Bottomley explains, the process of receiving and preparing audio, and then checking it for accuracy and quality, is time-consuming and typically involves a mesh of different software tools and project management tasks.

Doing quality assurance on one hour of audio can take two hours, he says. Multiply that by many hours of audio in multiple languages, and time savings can have a huge impact.

voxANN wants to simplify all of this with its own software tools, and then creating a marketplace-like platform that integrates with third-party apps and service providers to make the entire workflow seamless.

“On one side, there are LSPs–language service providers–that translate multimedia content. On the other side, we have what we refer to as creators. These are companies and individuals who are producing and owning content, they may have license it, and they want to distribute it,” Bottomley explains.

“voxANN sits in the middle, offering technical solutions to the industry and a democratised production process to the creators.”

The idea is that creators and media companies could upload content for localisation into different markets. Different service providers could then compete for this work.

The same platform could then be used to take that content and handle the localisation process with built-in tools and third-party software all in the same place, to make the workflows smooth and seamless. The creators can then then simply take the output away to use.

To begin with, the startup has developed a tool that handles the quality assurance process for translated audio that has been recorded in multiple languages.

Disparate audio files containing recordings by voice artists in different languages can all be synced up to a master script. This allows a professional to very quickly check everything is as it should be with both a voice-to-text read-out and the original audio available for playback across all the relevant languages.

The alternative workflow to voxANN’s software, called Align, involves lots of manual searching around inside audio files while looking at scripts, and trying to make sure everything is accurate without necessarily being able to understand all of the languages. A pain, in other words.

The story so far

voxANN has its roots in work Bottomley used to do with Dave Bilsborough, processing multilingual content for aircraft, for their employer. Amid the Covid pandemic travel slump, they were made redundant, with their work set to be outsourced to a third-party.

Bottomley says the pair took this opportunity to cannily set up a company to become that third party. Once they had that business up and running, they began to look at how tech could make their lives easier.

They came up with the vision of what they want voxANN to become (although it was called the slightly less snappy ‘Pre-recorded Announcement Production and Procurement Portal’ at the time), and they established their was demand from language service providers for such a product.

They quickly realised that if they expanded beyond just pre-recorded announcements, they could potentially build a whole new tech business for multi-language audio production.

So Bottomley and Bilsborough teamed up with Tom Crompton and Robin Tong of digital agency DiCE Creative to build these tools with the help of grant funding. While this was initially a commercial partnership, Crompton and Tong eventually came on board as technical co-founders to round out the voxANN team.

Align is set to launch in beta in the coming weeks. This follows a pilot programme with a group of language service providers, where the startup developed plans for their broader marketplace platform with feedback from potential customers. Align was the part that had the most immediate customer demand.

“We thought we could go ahead and build this whole big platform, but it would take far too long, and we needed to get something out into the market very quickly. So we set ourselves, a six- to eight-month target to build out this portion of the platform, which is automated quality assurance for multilingual audio, says Bilsborough.

“Initial pilot users were telling us ‘we're spending a lot of time checking audio, making sure that it's pronounced correctly, making sure that the voice artist has delivered it correctly and not given any kind of gaffe or added any extra phrases, and we have to listen through all of that in 10 or 15 languages without being native speakers, or outsourcing it to linguists, and we can't charge back to the client for that’.

“So we built Align with the specific purpose of solving that problem.”

Over time, voxANN plans to offer the ability to use AI-powered voices and translation as part of its marketplace.

Subscribe to Premium to read the rest.