The Growing Demand for Outsourced AI Data Annotation with Frank Prempeh of Corpshore Solutions

Frank Prempeh returns to the Outsource Accelerator Podcast, and this time the destination is not a country but a technology.

The CEO of Corpshore Solutions has built a career chasing emerging outsourcing frontiers, from the Dominican Republic to Uzbekistan, and his latest is AI data annotation. Through Corpshore AI and a new global platform, jwuma.com, he is betting on a market he says is growing, not shrinking.

The conversation digs into what annotation now involves, the widening spread of pay and quality, and the vast language opportunity ahead.

Corpshore’s pivot to AI

Frank describes a deliberate move from traditional outsourcing toward AI, prompted by client demand.

“We’re a Canadian-headquartered BPO, IT outsourcing, and AI managed services provider with operations in multiple parts of the globe.

Noticing the demand from clients, especially with the launch of ChatGPT, for AI data annotation, we pivoted to launching our Corpshore AI subdivision roughly about two years ago.”

The newest piece is a platform built to source annotators far beyond the usual markets.

“We recently launched our jwuma.com platform, which is our global AI annotation platform, where we try to welcome annotators from any parts of the globe, because there is demand for AI anywhere.”

What AI data annotation involves now

Annotation has moved well beyond labeling photos, Frank explains, as AI expands into new modalities.

“Prior to ChatGPT’s launch in late 2022, annotation was predominantly focused on text-based and image-based annotation.

Now you’re looking a lot at voice agents, voice-based data annotation, video annotation. Egocentric data annotation is a huge thing now because of the imminent rise of robotics.”

A widening spread of pay and quality

The economics vary enormously, from cents to hundreds of dollars an hour, depending on the complexity and the client.

“There are even roles that are being offered for $150 or $200 or even $300 an hour for certain type of roles, because of the demand in high-quality annotation.”

The buyers who grasp this, he says, are the biggest names in tech.

“The more reputable clients like Google or Amazon, Meta, they understand firsthand the importance of being able to ensure you get high-quality data.”

The language opportunity

One of the richest frontiers, Frank argues, is language, including tongues on the edge of disappearing.

“You have certain languages where you have three or five or 10 people in the entire world that can speak that language… The AI could be that canonical repository for that type of language.”

With annotators in more than 150 countries, Corpshore plans to build and license models for languages that mainstream AI has ignored, including under-represented ones like Arabic, with its non-Latin script and many dialects.

A growing, not shrinking, market

Far from being automated away, Frank sees annotation demand compounding as AI spreads into new products and markets.

“Right now, the global annotation market is approximately $4 billion per my latest research, but it’s gonna grow and expand over time, because there’s gonna be constant demand for these AI different products.”

He is careful, though, that data carries real obligations.

“A lot of this data as well is also under regulation, so you can’t just take any data and just sell it because a consumer’s offering top dollar for it.”

Quality over margin, and keeping humans real

Unusually for a volume business, Frank says Corpshore does not simply route work to the cheapest location.

“Are we gonna compromise the integrity of the contract and the integrity of our relationship with our client by opting for a higher margin location, as opposed to getting the best quality outcome out of it?

So we’re not necessarily led towards the best possible margin with the annotators.”

Keeping the workforce genuinely human is its own challenge, met with layered verification.

“We go through an extensive ID verification and camera liveness check to make sure that it’s a real person… We also use the major payment rails, so Stripe, to build a complete profile to confirm this person is indeed a human being and not a bot.”

BPO’s AI future

Frank is clear that traditional outsourcing is not disappearing, but evolving toward a human-plus-AI model.

“The best model is collaborative intelligence or augmented intelligence, where AI is automating a lot of the repetitive mundane tasks, but you still have an executive human decision maker.

The call center agent of the future is not gonna be somebody that’s just passively sitting there waiting for calls to come in.”

And on the core question of demand, he is emphatic.

“We’re incredibly bullish. Even internally as a company, we’ve seen way more demand now for offshore staffing and outsourcing services this year alone than we’ve experienced in previous years combined.”

For Frank, AI data annotation is not a threat to outsourcing but its next frontier: a growing, increasingly complex, globally sourced business where quality, trust, and language depth matter more than chasing the lowest wage.

To learn more about Corpshore’s AI data annotation services, visit corpshore.ai, and annotators looking for work can apply at jwuma.com.

And if you would like to learn more about what we do in the outsourcing space, send us an email at [email protected].