JapanHealthcare and life sciences2024 – ongoingSawai Group Holdings
Company-wide Gemini and a research RAG assistant for formulation work
Sawai's formulation researchers cut information gathering from up to 4 hours to about 15 minutes with an internal RAG assistant, the company said
Problem
New staff in formulation research spent one to several hours searching past reports for earlier cases and methods when facing a formulation problem.
Approach
The formulation research department built a RAG assistant on Azure OpenAI that draws on past study reports and shows reference links under each answer. Separately, the group introduced Gemini in a closed Google environment for all employees in April 2025. The communications team also used Gemini and NotebookLM for checking figures in drafts.
Results
Takeaway
A narrow assistant built on a department's own reports was used daily, while the general tool reached most staff.
Categories
- Size
- Large enterprise
- Industry
- Healthcare and life sciences
- Business function
- Internal work and knowledge managementIT and R&D
- Innovation type
- Process innovation
- AI technology
- Generative AI
- Adoption
- Off-the-shelf tool or platform
- Evidence
- Company's own claim
Sources
Source summaries
An interview with Sawai Group staff on its generative AI use. They described Gemini rolled out in April 2025 and a RAG assistant built by the formulation research department. They said information gathering fell from 1 to about 4 hours to about 15 minutes, and that Gemini usage ran at 70 to 80 percent.
Full details are in the original sources.