From a Campus Product to an AI Creation Portfolio
Mian Chen's verifiable starting point is not the founding of LiblibAI in 2023, but an attempt to build a Chinese GitHub with an older student while at Southeast University. A transcript of his interview says that the product failed after launch and that he then went to Tencent for an internship; a Southeast University news item records the Jihe team's startup plan winning a national Challenge Cup gold award in 2012. These two sources establish an early attempt at product building and entrepreneurship. They do not show that a registered company existed, nor do they establish the age of his “first startup.”
Before and after graduation, he interned at Tencent, then worked in product at 360 and Baidu Zhidao, at Didi, at Mobike and in the organization after Meituan acquired it, followed by product growth operations at Daily Fresh. He later joined ByteDance's education business and moved to global commercialization for Jianying CapCut after the double-reduction policy. This sequence comes from the interview transcript and public career materials; the precise employment dates still need verification. It shows that his accumulated capabilities span product, growth, organizational change, and overseas commercialization, and that he did not start his company with a complete set of resources from day one.
| Stage | Verifiable identity / organization | Evidence boundary |
|---|---|---|
| Early | Tencent intern | The interview transcript says he did not stay on |
| Product roles | 360, Baidu Zhidao, Didi, Mobike | Sequence follows the interview transcript; exact dates need verification |
| After acquisition | Stayed after Meituan acquired Mobike | This was not a separate job hop; see the interview transcript |
| Growth and overseas expansion | Product growth operations at Daily Fresh; ByteDance education business → global commercialization for Jianying CapCut | Public career materials and interview transcript; months need verification |
Cash Pressure and Product Choices
After founding LiblibAI in May 2023, Mian Chen recalled the early cash pressure in an interview with Zhang Xiaojun: at one point during 2023–2024, only about 4,000 yuan remained in the account. This is his own statement, not a company financial disclosure, but it explains why later funding and ARR cannot be read backward as evidence of an uninterrupted rise. Lovart moved from an internal test in May 2025 to a global launch in July, extending the product focus from a model and work community into a deliverable design agent.
Decision Principles
The experiences above suggest several tendencies in his product choices; the following is an analytical interpretation. The repeated actions we can observe are continuing to redefine the problem after a product failure, moving into globally demanded product contexts under external constraints such as the double-reduction policy, and breaking complex creative work into measurable credit consumption. They support an analysis that he adjusts products to trading and technical conditions, but do not prove that every pivot was driven by a single principle.
Evidence Boundary
The 2014 graduation and major, as well as the sequence of early roles, come from the interview transcript and are explicitly not independently replayed or verified here; the cash crisis is self-reported; Lovart's 2025 scale and the 2026 ARR, funding, and valuation come from media or company disclosures and have not been audited. Where public evidence was unavailable, I did not add specific months to employment histories, and I did not turn company ARR into Lovart revenue, personal wealth, or profit.