GIX Inc. Launches "AI-Informed Platform (AIP)" Service to Enhance Corporate Decision-Making

GIX Inc. (Headquarters: Minato-ku, Tokyo; CEO: Tomohiro Amino; hereinafter referred to as "the Company") is pleased to announce the launch of its new service, "AI-Informed Platform (AIP)," which combines generative AI and data infrastructure to support the enhancement of corporate decision-making processes.

AIP is a service designed to enable companies to make faster and more rational decisions by having AI agents cross-reference and analyze multiple data sources to support analysis and hypothesis generation. GIX aims to embed such AI-Informed decision-making processes, where "AI presents hypotheses and evidence, humans make judgments, and the field acts," into corporate organizations.

Background

Many companies have been investing in data infrastructure and BI tools as part of their DX promotion efforts.

However, there are many voices from the field, such as:

Data exists, but only a few specialized personnel can utilize it.

Dashboards have been prepared but are not fully utilized, and business departments continue to operate using Excel.

Generative AI has been tested, but it does not provide accurate enough answers to be useful in business.

The background to these challenges is not merely the presence or absence of tools, but the fact that "data is not designed and organized in a way that can answer the questions of AI and business departments."

For AI to understand business context and present meaningful hypotheses, it requires not only the整備 of data itself but also a semantic layer that defines the meaning and relationships of the data. We believe this data design philosophy is at the core of data utilization in the AI era.

AIP Concept

AIP aims to achieve a state where "AI interprets data, humans make judgments, and the organization acts."

For example, if a business manager asks in Japanese, "Why did sales in the Tokyo area decrease last month?" an AI agent will cross-reference multiple quantitative and qualitative data sources such as sales data, foot traffic data, weather data, policy information, meeting minutes, and store/sales reports, and present candidate factors, evidence, and points that require further verification.

Since the AI autonomously analyzes and presents hypotheses, the manager can make judgments based on those hypotheses and implement measures. Furthermore, the execution history and results of those judgments and measures are accumulated as data and become input for the AI's analysis in the next cycle.

The essence of AIP is not for AI to provide the final answer, but for AI to present hypotheses and evidence, humans to make judgments, and the organization to act, continuously repeating this cycle at the speed of the field.

GIX's Vision for an AI-Informed Decision-Making Foundation

Since its founding, GIX has focused not just on presenting analysis results, but on connecting data to business judgments and actions.

No matter how high-performance an AI is, it cannot generate hypotheses usable in business if the data is not organized according to the business context.

AIP is based on a three-layer design:

"Bronze": Holds raw data.

"Silver": Formats and integrates raw data into common, analyzable data.

"Gold": Optimizes data for specific purposes for business and decision-making.

While implementing this, we build a semantic layer that manages the meaning, definitions, and relationships of the data. This enables the AI to interpret data in the context of business operations and present evidence-based hypotheses.

For building this semantic layer, we utilize "Zokusei," an attribute definition framework independently developed by GIX. Zokusei is a crucial element for assigning business-related semantic information to data such as customers, products, stores, and policies, enabling AI to analyze data within the business context.

Our differentiated area lies in our ability to provide end-to-end support, not only in technical construction but also in business-oriented data design, analysis and hypothesis generation, integration into decision-making processes, and establishing PDCA cycles in the field. In particular, our long-standing know-how and assets are utilized in compressing vast amounts of data information and aggregating/transforming it into data suitable for analysis.

AIP is a customized service designed and built according to each company's data infrastructure, business challenges, and decision-making processes. The scope of implementation and provided content are determined through individual consultations.

We begin with an "AIP Suitability Diagnosis" as a current assessment. This diagnosis organizes the current situation from four perspectives: ① Data Infrastructure (Technology), ② Data Utilization (Business), ③ Systems, and ④ Organization/Processes, and creates a transformation roadmap towards an AI-Informed organization.

AIP Usage Example

To mark the launch of this service, we conducted a demo analysis using a fictional Italian restaurant chain as a case study.

In this demo, we analyzed data across sales, weather, foot traffic, local events, internal policies, meeting minutes, and store daily reports to understand "why the Kichijoji store's March sales increased by +20.6% year-on-year." The results presented candidate factors and evidence, such as the overlap of a cherry blossom viewing event, a takeout promotion, and improvements in store operations.

In this demo, the AI does not determine causality but presents a set of hypotheses, evidence, and additional verification items based on the data. The approach is not to blindly accept the AI's output, but to use it as material for humans to make better judgments – this is GIX's AI-Informed approach.

Future Development

Through the provision of AIP, GIX will continue to support companies in end-to-end data and AI utilization.

We will respond to a wide range of needs, from new data infrastructure construction, diagnosis and improvement of existing infrastructure, semantic layer construction, analysis and hypothesis generation using generative AI, to support for establishing decision-making processes.

You can start by consulting us for the "AIP Suitability Diagnosis" to organize the current state of your data, systems, and organization.

Inquiries and Consultations

About GIX Inc. https://www.gixo.jp/

GIX is a "data-informed" promotion company founded by strategic consultants and analytics experts. We support client companies' "customer understanding" by maximizing data utilization and solve management challenges for business growth. We aim to transform the world using technology and methodologies, enabling humans to think logically and make rational judgments based on data, rather than blindly relying on it.

Company Name: GIX Inc.

Representative: Tomohiro Amino, CEO

Location: Mita Kokusai Building 11F, 1-4-28 Mita, Minato-ku, Tokyo

Business Activities: Data-informed business using analytics

Consulting services utilizing data, and research and development of tools

Provision of various services using the above tools