← Cases
ChinaLogistics and transport2024 – ongoing

S.F. Express

Fengyu assistant for customer-service agents

S.F. Express reported 30% less after-call work with Fengyu conversation summaries

WorkingCompany's own claim

Problem

Customer-service agents needed to record inquiries after conversations. Subsequent agents needed those records to handle follow-up calls.

Approach

S.F. Express built Fengyu, a logistics-focused large language model. It supplied suggested answers and extracted information for service tickets. After conversations, it recorded inquiry details and generated summaries.

Results

30%reduction in average post-conversation processing time, reported by S.F. ExpressCompany's own claim[1]
>95%customer-service summary accuracy, reported by S.F. ExpressCompany's own claim[1]

Takeaway

The reported time saving concerned work after customer conversations. S.F. Express did not disclose a reduction in total handling time.

Categories

Business function
Customer service
Innovation type
Process innovation
AI technology
Generative AI
Evidence
Company's own claim

Sources

Source summaries

[1]Nanfang Plus14 Sep 2024

Nanfang Plus described Fengyu's use in customer service. It relayed S.F. Express's figures of summary accuracy above 95% and 30% less post-conversation processing time.

Full details are in the original sources.

Added 1 Nov 2026 · Updated 1 Nov 2026
Open case page ↗