• Appen, the Australian AI firm that once helped train AI models for tech behemoths including Microsoft, Nvidia and Google, has lost its chief revenue officer and chief marketing officer.
• The departures took place last week, according to an internal memo viewed by CNBC.
• Alphabet in January said that it would cut all contractual ties with Appen.

Appen, the embattled artificial intelligence firm that once helped train AI models for tech giants including Microsoft, Nvidia and Google, has lost its executives in charge of revenue and marketing.

Andrew Ettinger, who was the Australian company's chief revenue officer, and Alicia Hale, who was marketing chief, stepped down from their roles last week, according to an internal memo viewed by CNBC. Both executives joined the company last year.

"Strengthening our sales and marketing function remains a top priority for the business," CEO Ryan Kolln wrote in the memo that was shared with CNBC. "There is no change to our strategy to grow revenue from existing and new customers."

The departures follow Alphabet's announcement in January that it was cutting all contractual ties with Appen, which once helped train Google's chatbot and other AI products. Two weeks after that decision, Appen CEO Armughan Ahmad left after just 12 months on the job.

Although generative AI is booming, Appen, once an industry darling, has been losing out on business as tech companies spend billions of dollars training their own large language models (LLMs) or building atop the leading AI platforms. They're all pursuing a market that's predicted to top $1 trillion in revenue within a decade.

Despite Appen's once-enviable client list and its nearly 30-year history, revenue dropped 30% in 2023, after declining 13% a year earlier. The company attributed the decline in part to "challenging external operating and macro conditions."

Former employees told CNBC last year that the company's struggle to pivot to generative AI reflected years of weak quality controls and a disjointed organizational structure.

The latest memo also mentioned that the company's vice president of sales and vice president of global solutions will now report directly to Kolln, who wrote that the company is "targeting customers that are currently spending on data services."

In the past, five customers — Microsoft, Apple, Meta, Google and Amazon — accounted for 80% of Appen's revenue, and the company used its platform of about 1 million freelance workers in more than 170 countries to train some of the world's leading AI systems.

After a "strategic review process," Alphabet notified Appen in January of the termination, which went into effect March 19, according to a filing from Appen. The company said at the time it had "no prior knowledge of Google's decision to terminate the contract." In 2023, revenue from work with Alphabet totaled $82.8 million of Appen's $273 million in sales for the year, according to a January filing.

In August 2020, Appen's shares peaked at AU$42.44 ($27.08) on the Australian Securities Exchange, sending its market cap to the equivalent of $4.3 billion. The company has since lost 99% of its value.

Appen's past work for tech companies has been on projects like evaluating the relevance of search results, helping AI assistants understand requests in different accents, categorizing e-commerce images using AI and building out map locations of electric vehicle charging stations, according to public information and interviews conducted by CNBC.

The LLMs of today that are behind OpenAI's ChatGPT and Google's Gemini are scouring the digital universe to provide sophisticated answers and advanced images in response to simple text queries. Companies are spending far more on processors from Nvidia and less on external AI training from companies like Appen.

"I'm highly focused on supporting our sales team so they can be as effective as possible," Kolln wrote in the memo. "To achieve this, we need to equip them with the content and messaging that differentiate Appen vs our competitors."

Appen didn't immediately respond to a request for comment.

WATCH: 'We're a decade away from' from solving AI