Gabe Greenberg wants to add 50 software engineers to his team within a week. The CEO and founder of G2i, a Florida-based engineering talent platform, posted the call on X earlier this week, offering between $100 and $200 per hour for fully remote work available in more than 150 countries.

The work involves helping train frontier AI models. Greenberg's post stated that G2i works "with frontier labs to help them train models." In practice, this means reinforcement learning from human feedback, or RLHF, where experienced engineers evaluate, rank, and critique code that AI systems generate.

What the Job Actually Involves

According to G2i's job listings, engineers on these projects don't build products in the traditional sense. Instead, they compare multiple code snippets and explain which is better and why, repair and refactor AI-generated code for correctness and efficiency, write test cases, and inject feedback into the RLHF pipeline. As G2i describes it in one job posting: "the model learns to propose, critique, and improve code the way you do."

The company requires at least three to four years of professional software engineering experience, depending on the specific role. Strong code review instincts and the ability to spot logic errors, performance issues, and security problems quickly are essential. Prior AI or machine learning experience is not required. G2i says it will teach engineers the RLHF specifics.

G2i accepts applicants from the United States, Canada, most of Latin America and Europe, and some countries in Africa and Asia. The work is classified as 1099 contractor roles for US-based workers, which means it's not compatible with visa statuses that require W-2 employment or guaranteed hours.

The Company Behind the Hiring Blitz

G2i has operated since 2016 as a video-based hiring platform for software engineers, initially focused on React and React Native developers. The company claims to have placed over 2,300 engineers and paid out more than $28 million to developers. Greenberg founded the company after attending Florida Atlantic University and has built it without venture capital backing, according to interviews he's given on tech podcasts.

The pivot toward AI training work represents a significant expansion of G2i's business model. The company now lists over 200 open positions for AI training data work across various programming languages including Python, JavaScript, Java, Rust, Swift, and C#. It handles employment compliance across its stated 150+ countries, a capability that positions it well for the globally distributed nature of AI-related contract work.

The Pay Question

The $100 to $200 hourly rate Greenberg advertised is at the higher end of what G2i typically offers for this type of work. Other G2i job postings for AI training roles show rates ranging from $30 to $70 per hour. Market data for freelance LLM developers shows wide variance based on experience and specialization. The work is project-based and flexible, with most engineers putting in between 15 and 40 hours per week.

G2i's onboarding process is designed to be completed within a week. Candidates go through an asynchronous RLHF code review, receive platform credentials, complete a training course, and perform a production-level task that serves as the final interview. The entire process takes roughly 3.5 hours of active work.

The rapid hiring push reflects the intense demand for human expertise in AI development. While layoffs continue across parts of the tech sector, the infrastructure work of training AI systems requires a growing army of contractors who can evaluate whether generated code actually works. G2i is positioning itself as a broker for that labor.