JapanTravel and aviation2023 – ongoingJR West Customer Relations
Generative AI summaries of customer calls
JR West Customer Relations cut assistance-request summary time by about 54% in an ELYZA pilot
Problem
Operators manually summarized calls for shared records. Writing these summaries took substantial time, and quality varied between operators.
Approach
The company and ELYZA tested generated call summaries from May to August 2023. ELYZA used GPT-3.5 and GPT-4 according to inquiry type. Wider operation began in September after the pilot.
Results
Takeaway
Savings varied by call type. ELYZA attributed the largest reduction to assistance requests with standardized questions.
Categories
- Size
- Large enterprise
- Industry
- Travel and aviation
- Business function
- Customer service
- Innovation type
- Process innovation
- AI technology
- Generative AI
- Adoption
- Partner co-development
- Evidence
- Vendor case study
Sources
Source summaries
ITmedia reported ELYZA's pilot results and subsequent rollout. Assistance-request summary time fell by about 54%; other call categories recorded smaller reductions.
Full details are in the original sources.
