Australia-headquartered Appen published results for the first half of 2018. Appen is listed on the Sydney stock exchange and trades under the ticker APX. Appen has two business lines: a Language Resources Division that provides datasets (audio, text, image and video) for training AI engines, and a Content Relevance Division that helps clients train AI driven products (mainly search engines) via human evaluation and feedback.

The company employs 394 full-time staff and generates most of its revenues in the United States and in USD. Much of Appen’s work is farmed out to a global network of freelancers. The company says it currently maintains a so-called “on-demand global crowd” of around 1 million people.

In late 2017, Appen acquired Silicon Valley-based Leapforce to build out its search capabilities for the Content Relevance business. As a result, revenue were up 106% to AUD 152.8m (USD 112.3m) in the first half 2018. Organic revenue growth was still an impressive 47%. EBITDA doubled to AUD 25.6m (USD 18.8m).

Much of the growth was driven by the Content Relevance division, whose main service is providing human evaluation of search engine results. Appen works with tens of thousands of human, in-country evaluators who judge the local relevance of search engine results.

In late 2016, Appen CEO Mark Brayan told Slator that the division gets “tens of thousands of queries and we farm them out to the crowd. It’s a query pair: a query and a URL. We look at the query and we look at the URL and, yes, there are certain guidelines, but it’s really up to the human evaluator whether that’s a good response to that query”, Brayan said at the time. The unit’s largest clients are likely Google and Microsoft’s Bing, although the company does not name its clients publicly.

Content Relevance revenues grew to AUD 131.2m (USD 96.4m) in H1 2018 from AUD 53.3m (USD 39.2m) in the same period in 2017. Even excluding the Leapforce acquisition, organic revenue was up 64% and EBITDA up 88%.

Revenues at the Language Resources division grew by a much more moderate 4%. EBITDA was down by more than half from AUD 7.5m (USD 5.5m) to AUD 3.4m (USD 2.5m). Appen says the slowdown was due to timing issues and not underlying weakness in the business.

The Language Resources division creates audio databases for training speech recognition engines. “That includes capturing audio on the field to make sure we get the right accents, languages and, in some cases, the right situations and acoustic qualities,” Brayan explained in the 2016 interview.

Appen said as of end of July 2018, year-to-date revenues plus orders in hand for 2018 already totalled around AUD 250m (USD 183.7m), guiding for a strong revenue figure for the full year. The company forecast full-year 2018 EBITDA to be in the range between AUD 54m (USD 39.7m) and AUD 59m (USD 43.4m).

On the markets, Appen continues its phenomenal run. Since going public in January 2015, shares in Appen are up an eye-popping 2,700%. The company’s current market capitalization stands at AUD 1.6 billion (USD 1.2bn).