(Yicai) June 26 -- DeepSeek, which has quickly become one of the world’s most valuable artificial intelligence startups at a reported CNY400 billion (USD58.8 billion), plans to at least double its headcount in a bid to spur development of artificial general intelligence, which can match or surpass human cognitive abilities.

DeepSeek intends to “at least double the size of every department,” the Hangzhou-based company  announced yesterday. “Humanity is on the eve of AGI,” it said, pointing out that new hires will have the opportunity to take part in the most critical and important tasks, becoming a key “driving force” in advancing AGI development.

Set up in 2023, DeepSeek burst onto the global AI scene in January last year with the public release of its relatively low-cost model, DeepSeek-R1. Silicon Valley venture capitalist Marc Andreessen famously described it as an “AI Sputnik moment.” It reportedly raised CNY51 billion (USD7.5 billion) in a Series A funding round earlier this month from backers such as founder Liang Wenfeng, Tencent Holdings, NetEase, JD.Com, and other Chinese tech giants, at a valuation of CNY400 billion.

DeepSeek is hiring in 33 positions across seven major categories, including full-stack development and algorithms, AI core systems development, operations and maintenance, product, model data strategy, deep learning research, and corporate functions. The roles are located in Beijing and Hangzhou.

Recruitment is still focused on full-stack development and algorithms, with eight positions, including roles for the Agent Harness team, which published its own hiring notice last month. That team is devoted to developing code-centric AI agents and benchmarks its work against Anthropic’s coding tool Claude Code, Senior Researcher Chen Deli said previously.

DeepSeek has also opened a unique position for AI interdisciplinary talent, targeting individuals with "extraordinary abilities who aspire to contribute to the creation and development of AGI," with no restrictions on professional background.

Editors: Dou Shicong, Martin Kadiev