Back to case studies
AIIT & SaaS
Internal knowledge chatbot (RAG)
Unified internal docs in a vector DB and built an LLM Q&A chatbot, available inside Teams and Slack.
- Client
- SaaS company (Japan)
- Industry
- IT & SaaS
- Market
- Japanese client
- Duration
- 4 months
- Team
- 3 (AI engineer, full-stack engineer, PM)
Key metrics
-25%
First-line workload
Up
Self-resolution
Faster
New-hire ramp-up
Challenge
Knowledge was scattered across departments, so answering questions and finding information was slow.
Approach
Consolidated internal documents and built a vector DB + LLM chatbot with cited answers, embedded in the Teams/Slack tools staff already use.
Scope of work
- Document collection & cleaning
- RAG pipeline with citations
- Teams / Slack integration
- Answer-quality evaluation
Results
Self-service resolution of common questions rose, first-line workload fell 20–30%, and new hires ramped up faster.