Gao Hong worked as a frontend engineer at Shopee. Around April-May last year, her boss already sensed the risk and started pushing everyone to work on AI agent projects.
Over more than a year, she participated in everything from project inception to launch, from collecting user data to establishing benchmarks and process metrics. But even with consistently good performance, she was laid off.
Her boss told her: "It's hard to 'climb out' from frontend anymore. Even without AI, the career advancement path had already narrowed. But with AI, there's a new opportunity—everyone's starting on the same new road. As long as you understand a little more than others, you might have an advantage in the job market."
A few years ago, when interviewing for a frontend position at ByteDance, she often felt her technical knowledge could be 'questioned.' But when she started targeting agent engineer roles, she discovered something surprising: "When I interviewed as an agent engineer, the interviewer often didn't know more than me."
She's been communicating with startup companies recently. They're building agents for gaming, finance, and other vertical industries, discussing specific business scenarios: how to design workflows, how to judge results, where things might go wrong. Those interviews felt more like collaborative explorations.
"Frontend work久了, knowledge boundaries felt familiar, even somewhat tired. But in AI and agent fields, there's still so much worth exploring. It even helped me rediscover my passion for technology," she said. "Many times, opportunities and seemingly pessimistic situations coexist."