Fuke AI — A Gen Z-founded startup achieving an 80% AI customer service response rate in e-commerce

Origin

The founder, a member of Gen Z, earned 10 million RMB through other business ventures during college before choosing to enter the AI customer service sector. The core motivation was the high cost of human customer service for e-commerce merchants, significant lost sales due to lack of overnight coverage, and insufficient capacity during peak promotional periods. The team aimed to use vertical AI Agents to handle automated responses, product recommendations, and after-sales processing, focusing on improving response and conversion rates during off-hours and peak sales. After receiving investment from Alibaba, the company entered a period of rapid expansion, but by 2026, the leap in general-purpose LLM capabilities compressed the survival space for vertical agents, forcing the team to seek new moats.

Milestones

Turning Points

• The founder entered the AI sector with 10 million RMB earned in college; early business intuition became an advantage in customer acquisition.
• Achieving an 80% AI response rate became the key metric to win Alibaba's investment, triggering a period of rapid expansion.
• The 2026 leap in general-purpose LLM Agent capabilities commoditized vertical AI customer service, leading to client churn.
• The founder publicly reflected on how vanity and over-packaging caused the team to lose direction, forcing a downsizing and a pivot to deep vertical scenarios.

Failures & Pitfalls

• Due to vanity and a performative culture, the founder over-packaged technical capabilities early on, ignoring real merchant feedback and resulting in unused product features.
• After the leap in general-purpose model capabilities, reliance on vertical prompt engineering and knowledge base management ceased to be a moat, and the 80% response rate lost its scarcity.
• The team expanded too quickly after funding, wasting resources on functions that would eventually be covered for free by general-purpose models.
• In a podcast interview, the founder admitted to having been at the 'peak of stupidity,' severely underestimating the difficulty of AI entrepreneurship and the threat posed by general-purpose models.

Core Data

• Founder's earnings during college：10 million RMB (based on public information, not independently verified)
• Funding event：Received investment from Alibaba; specific round and amount undisclosed (based on public information, not independently verified)
• Team size：Not disclosed (based on public information, not independently verified)
• Monthly active merchants：Not disclosed (based on public information, not independently verified)
• AI customer service response rate：80% (based on public information, not independently verified)

Competitors / Peers

The vertical AI customer service Agent market is highly competitive. Domestic rivals include RPA+AI customer service vendors like Laiye (ai-agent-rpa-cs-laiye), as well as AI outbound and customer service Agent companies like Silicon Intelligence and Yizhi Intelligence. The biggest competitive threat in 2026 comes from the native Agent capabilities of general-purpose LLM providers, such as ByteDance's Doubao, Alibaba's Tongyi Qianwen, and Baidu's Ernie Bot, which directly cover e-commerce scenarios, pressuring vertical vendors to be absorbed or replaced by platform ecosystems. Additionally, a group of gray-market service providers (ai-customer-service-outsourcing) are capturing the low-end merchant market with low-cost outsourcing models, further compressing the pricing power of vertical Agent startups.