From Side Hustle To Seven Figures: How Prompt Training Turned Into a Lucrative AI Service

Mark Kashef launched Prompt Advisers within days of ChatGPT’s debut and has since scaled the operation into a business expected to pull in between $1.5 million and $2 million this year. The agency splits its revenue roughly 7/10 from consulting and training and 3/10 from bespoke development, serving a mix of entrepreneurs and small companies eager to fold generative AI into everyday work. Kashef says the firm’s rise began with simple listings on a freelance marketplace and grew into a steady flow of clients after he built an audience online.

Kashef traces his edge to a 2019 master’s in management and artificial intelligence, where he encountered what would become his specialty: instructing models to produce useful outputs. In late 2022, a few days after ChatGPT hit public attention, he put up an offer on Fiverr to craft prompts for clients. Early demand was modest and he charged only a few dollars while testing the market. Within months he adjusted prices several times — from single digits to tens, then to hundreds of dollars for more thorough prompt packages — as requests ballooned.

Visibility on the freelance platform helped. Kashef’s gigs rose to the top for the term “prompt engineering,” drawing about 50,000 impressions a month on that search alone. Platform-driven publicity and a handful of press placements broadened the firm’s reach, enabling it to expand services beyond prompt writing into workflow automation and product-ready AI tooling. All the while he kept a day job as a data science manager, fitting business work into early mornings and late evenings until he felt confident the enterprise could shoulder payroll after he added a partner and several staffers.

A push into video content roughly 18 months ago accelerated inbound interest. Regular tutorials and case studies on YouTube amplified credibility and funneled new clients into the consulting pipeline, Kashef says. He admits he could have left his salaried role sooner but chose caution: when employees depend on you, the calculus changes. “I didn’t want to trade my steady paycheck for uncertainty until I could prove the demand would hold,” he told me in a recent interview.

The market backdrop favors specialists. A mid-2026 Fiverr trends report shows demand for developers working with a popular AI coding assistant has jumped by roughly 10 times in the past half-year, as businesses prioritize automation and faster product cycles. “Companies aren’t buying AI for novelty; they want repeatable improvements in productivity,” said Lena Ortiz, a consultant who helps small firms adopt automation. “That’s why training people to ask the right questions of these tools can be worth as much as custom-building systems for them.”

For consumers and small business owners, Kashef’s path offers a practical lesson: start lean, demonstrate value, then scale services as willingness to pay increases. He now packages training, templates and tailored builds, and recommends clients focus on three concise principles he teaches repeatedly — clarify the goal, set constraints, and iterate quickly — so teams can get useful results without overengineering. As more organizations seek to automate routine work, entrepreneurs who teach those skills may find a willing market.