Basic Inc. Presents at Microsoft Japan's Generative AI Event for the Pharmaceutical Industry, 'Columbus Day for Pharmaceutical Industry'

Basic Inc. (Headquarters: Shinjuku-ku, Tokyo; Representative Director: Katsuhiko Okuda; hereinafter "Basic"), a company supporting corporate DX promotion and generative AI utilization, is pleased to announce its participation as a speaker at "Columbus Day for Pharmaceutical Industry," an event hosted by Microsoft Japan Co., Ltd. (Headquarters: Minato-ku, Tokyo; hereinafter "Microsoft Japan") on February 27, 2026.

■ What is Columbus Day for Pharmaceutical Industry?

This event is held regularly as part of the generative AI business support program provided by Microsoft Japan. This session was held with the aim of promoting the use of generative AI and accelerating innovation in the healthcare and pharmaceutical industries.

As a partner company of Microsoft Japan, Basic participated as a speaker and gave a lecture on the theme of "Knowledge Utilization in Formulation Research," presenting specific problem-solving cases using the latest multi-agent technology.

■ Case Study: Construction of the Multi-Agent Platform "Sawai Agents" for Sawai Pharmaceutical Co., Ltd.

In the lecture, we introduced the construction of "Sawai Agents," a multi-agent platform for Sawai Pharmaceutical Co., Ltd., which we supported under the theme of knowledge utilization in formulation research.

In the field of formulation research, various challenges exist, such as "searching accumulated data" and "creating reports." In this project, we developed a platform to centrally manage task-specific agents on the Microsoft Azure (hereinafter "Azure") AI infrastructure, aiming to create an "AI that can be consulted like a veteran researcher" by making the most of vast amounts of data.

During the session, we explained specific approaches to three challenges we faced: "materializing specialized knowledge," "complex document analysis," and "adoption in the field."

We introduced methods that achieved answer quality equivalent to that of a veteran researcher by linking specialized agents for structural analysis, chunking, and other tasks to realize a high-precision knowledge base, combined with iterative exhaustive searches. Furthermore, we reported on the results of promoting knowledge sharing through human-AI collaboration by reflecting feedback from the field through a high-speed improvement cycle.

■ About the AI Agent Platform "Agent ARK"

At the end of the lecture, we introduced "Agent ARK," the multi-agent platform that serves as the foundation for this case study.

"Agent ARK" is an AI agent development platform that consolidates our development expertise. By incorporating a customer's unique know-how, it is possible to quickly build and launch agents specialized for their specific business operations.

Currently, we are offering a free "Agent ARK Trial" where you can try a specialized agent sample that has incorporated publicly available formulation data, allowing you to get a concrete image of knowledge utilization without incurring development costs.

Please feel free to contact us for inquiries regarding the Agent ARK Trial or consultations regarding the introduction of AI agents.