Latest / The Side Hustle Career Podcast with Fexingo: Second Income, Freelancing While Employed / How a Customer Support Rep Built a 12K Side Hustle in AI Training
Transcript
- Lucas: So you know how everyone talks about AI taking jobs? Well, here's a story about someone who used AI to build a side hustle. Maria worked as a customer support specialist at a SaaS company in Austin, making around $45,000 a year. Two years later, she's earning an extra $12,000 a month training the very AI models that everyone's worried about. Luna: That's a pretty dramatic jump. How did she even get into that? Lucas: It started when she saw a post on Reddit about data annotation work — basically labeling images or text to help train machine learning models. She applied to a platform called Remotasks, which is owned by a company called Scale AI. The first project she got was labeling street-level imagery for self-driving car training. Luna: So she's helping teach cars to see. That's a far cry from resetting passwords and handling support tickets. Lucas: Exactly. And the pay was decent — about $15 an hour to start. But Maria quickly realized that if she could specialize in a more complex type of annotation, she could earn more. She started taking courses and certifications offered by the platform on how to do lidar annotation and 3D bounding boxes. Within three months, she was making $30 an hour. Luna: So she doubled her hourly rate just by upskilling within the same platform. That's clever. Lucas: Right. But the real breakthrough came when she got invited to a private project for a major AI research lab. That project required annotating conversational data — teaching the model how to understand sarcasm, context, and nuance. The rate was $40 an hour, and she was allowed to work up to 40 hours a week. That's $6,400 a month. Luna: But you said she's making $12,000 now. How does she get to double that? Lucas: She started managing a small team. The platform allows top annotators to become reviewers and then project managers. Maria now manages a team of five other annotators. She reviews their work, handles quality control, and earns a percentage of their output. She also takes on multiple projects at once, balancing her schedule carefully. She treats it like a second full-time job, but on her own time. Luna: That's the key — she didn't just work harder, she changed her role. She went from a worker to a manager. Lucas: And honestly, stories like Maria's are exactly why we do this show. We want to show that with a little strategy, you can create real income outside your day job. Luna: It's inspiring. And you know, we keep this show ad-free on purpose. We think the advice should come without commercial interruptions. If listeners want to support that model, there's a simple way: buy me a coffee dot com slash fexingo. Lucas: Yeah, that's it. No pressure, just an option for those who find value here. So Maria's team management move — how did she actually go about finding those first team members? Luna: I imagine she had to vet people pretty carefully. Annotation quality matters a lot. Lucas: She did. She started by recruiting from the platform's community forums. She looked for people who had high accuracy scores and positive feedback. Then she ran a small test batch with each candidate before adding them to her team. She also had to train them on the specific guidelines for each project. Luna: So she became a mini-manager, handling training, quality, and payroll essentially. How does she bill clients for the team's work? Lucas: The platform handles most of the billing. Maria gets a percentage — typically around 10 to 15 percent of the team's gross earnings. So with five people each producing about $6,000 a month, that's $30,000 in output. Her cut adds up to $3,000 to $4,500. Plus her own active annotation work, she hits $12,000. Luna: That math works. But what about her day job? She's still doing customer support, right? Lucas: Yes, and that's the challenge. Maria works her day job from 9 to 5, then spends about three hours on the side hustle in the evenings, plus a few hours on weekends. She dedicates weekends to reviewing the team's work and handling administrative stuff. She says it's exhausting but worth it — she's saving for a down payment on a house. Luna: I can see why she wouldn't quit her day job yet. The side hustle income isn't guaranteed. The projects come and go. Lucas: Exactly. And Maria has experienced dry spells. Last year, one of her main projects ended unexpectedly, and her income dropped to $4,000 for two months. That's why she diversifies — she's on three different platforms now: Remotasks, Lionbridge, and Appen. She also does direct contracting with a couple of AI startups. Luna: So she's essentially running a small agency. Do you think this model could work for someone who doesn't have a customer support background? Lucas: Absolutely. Data annotation doesn't require a specific degree. What matters is attention to detail, reading comprehension, and the ability to follow complex guidelines. Maria says her customer support experience actually helped — she was used to following scripts and procedures. But she's seen people from all backgrounds succeed: teachers, retail workers, even a former truck driver. Luna: Speaking of which, our truck driver episode was a hit. So there's clearly a pattern here — find a niche within a growing industry, upskill, and then leverage team dynamics. Lucas: Right. And the timing is important. AI training work is booming, but it won't last forever. Once models are trained, the need for human annotators might shrink. Maria is already thinking about her next move — she's learning to code so she can transition into machine learning operations. Luna: Smart. So she's using the side hustle to fund a career pivot. That's a whole other level. Lucas: Yeah. And I think that's the real takeaway. The money is great, but the skills and network you build — managing a remote team, understanding AI pipelines, getting referrals from clients — those open doors you wouldn't expect. Luna: So for someone listening right now who's curious, what's the first step you'd recommend? Lucas: Go to Remotasks or Appen and take their qualification tests. They're free. See if you can pass the basic image annotation test. If you do, start with a few hours a week. Don't quit your job. Treat it like a learning experiment. The goal is to find out if you like the work and if you can earn consistently. Luna: And then after a few months, evaluate whether to specialize or scale. Lucas: Exactly. Maria's story shows that a side hustle in AI training is accessible if you're willing to learn and be patient. It's not get rich quick, but it can be a serious income stream. Luna: I think we've given listeners a concrete case to think about. Thanks, Lucas. Lucas: Thank you. And if you have your own side hustle story, we'd love to hear from you. Until next time.