JapanTechnology and telecom2025 – ongoingServerworks
Generative AI trainer bot for engineer training
Serverworks said a generative AI trainer bot cut the workload of customer-role trainers by about 40%
Problem
Serverworks wanted to hire more engineers, but training capacity limited it. In simulated AWS projects, trainers played the customer and sales roles and answered trainees' questions by hand.
Approach
Serverworks built a generative AI trainer bot. A trainee posts a question in Backlog and the bot drafts a customer-style reply, then notifies the trainer in Slack. The trainer either approves the draft with one click or edits it before it posts back to Backlog. The bot keeps context from past project-channel conversations and uses the existing training evaluation rubrics as reference.
Results
Takeaway
A trainer still reviews every reply. The saving came from drafting, not from removing the human check.
Categories
- Industry
- Technology and telecom
- Business function
- Human resources
- Innovation type
- Process innovation
- AI technology
- Generative AI
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
The company's engineering blog described how it built a bot that drafts customer-role replies in simulated AWS training projects. It explained the Backlog and Slack workflow in which trainers approve or edit drafts. It stated that workload for the customer-role staff fell by about 40%.
Full details are in the original sources.
