~Toward Making Japanese Aquaculture a 'Predictable' Industry with AI~ Keio Corporation and Nōka.AI Launch Okutama Yamame Trout Farming Optimization Project Utilizing Japan's First 'Biological Inference AI'

Keio Corporation (Head Office: Tama City, Tokyo; President: Satoshi Tsumura; hereinafter "Keio") and Nōka.AI, Inc. (Head Office: California, USA; Director: Naka Honda; hereinafter "Nōka.AI") will jointly launch the "Okutama Yamame Trout Farming Optimization Project" (hereinafter the "Project") starting Wednesday, July 8.

The Project is a collaborative research initiative aimed at promoting the growth of Okutama yamame trout, improving feed efficiency, and stabilizing quality, marking a pioneering effort that utilizes Japan's first "biological inference AI."

Aquaponics Business: A New Venture Originating from Keio Corporation Employees

Keio Corporation launched an aquaponics business in March 2026, utilizing underused facilities at Keio Plaza Hotel Hachioji, as a commercialization project from its open innovation program "My turn." This program fosters new businesses through co-creation with external companies, including startups, based on the ideas and aspirations of its employees. The project produces Okutama yamame trout and watercress within the former hot spring facility for pool users, under the concept of "the ultimate local production for local consumption: grown at the hotel, savored at the hotel."

Among these, Okutama yamame trout is a brand fish from the Okutama area, attracting attention from chefs and gourmands for its rarity as a raw river fish and its excellent flesh quality. To maximize the value of this ingredient and deliver it continuously to restaurants seeking high-value products, Keio recognized the importance of a system that scientifically understands the growth process in aquaculture, supporting production efficiency and quality stabilization. Establishing such a process had been a challenge.

<<Scene from the aquaponics business conducted at Keio Plaza Hotel Hachioji>>

Transforming Japan's Fisheries Industry with Nōka.AI's "Biological Inference AI"

Nōka.AI is a deep tech company born from the goal of fundamentally optimizing food production through AI-driven biological data analysis, addressing the challenges of rapidly increasing protein demand and the lack of adoption of productivity-enhancing technologies in the fisheries industry.

Japan's fisheries industry is a vital sector worth trillions of yen, and fishing is indispensable in supporting Japanese food culture. However, in aquaculture operations that support this industry, feed design and growth management often rely on years of experience and intuition. Fundamental questions such as "Why do fish grow?" "What influences metabolism?" and "How are nutrients utilized at the cellular level?" have remained largely un-systematized to this day.

To solve these challenges with science and technology, an AI specialized for industrial applications is required. Nōka.AI has therefore developed "biological inference AI," which reasons based on an understanding of biological mechanisms, replacing experience and intuition. By integrating vast amounts of information such as fish genes, blood, and muscle composition (multi-omics data *1), which it learns to understand biology, and various data including genome-scale metabolic models *2, transcriptomics, and proteomics *3, it becomes possible to analyze fish cell function, nutrient response in metabolic networks, and even the impact of environmental conditions on biological functions at the molecular level.

Nōka.AI aims to redefine food production itself through biological inference AI, which reasons based on the molecular-level logic of biological systems, extending beyond single projects or fish species. By enabling AI to understand organisms at the cellular level, producers can achieve higher productivity and optimize quality.

This project came to fruition from the convergence of Nōka.AI's achievement in developing AI that can scientifically answer the fundamental question of "Why do fish grow?" and Keio Corporation's desire to scientifically grasp the factors influencing growth to stably deliver Okutama yamame trout as a high-value ingredient.

*1 Multi-omics data: Refers to the integrated combination of data from multiple hierarchical levels that comprehensively capture life phenomena at the molecular level, such as genomics (genes), transcriptomics (gene expression), proteomics (proteins), and metabolomics (metabolites) of an organism.

*2 Genome-scale metabolic model: A mathematical model that reproduces the entire network of chemical reactions (metabolism) occurring within cells based on almost all genetic information of an organism. It can predict how nutrients are taken in and converted into energy and body components through computation.

*3 Transcriptomics/Proteomics: Transcriptomics refers to the overall picture of genes (mRNA) active within a cell, while proteomics refers to the overall picture of proteins present in a cell. Both are methods for comprehensively analyzing the extent to which specific genes and proteins are functioning.

About the Okutama Yamame Trout Farming Optimization Project

In this project, we will scientifically address the three core challenges of aquaculture for Okutama yamame trout: "What to feed them (feed)," "In what environment to raise them (farming environment)," and "When to harvest them (quality prediction)," using "biological inference AI."

First, for "feed," Nōka.AI will analyze fish metabolic data to identify nutrients that truly contribute to growth and quality, verifying feed design that does not rely on empirical rules. For the "farming environment," we will analyze the interaction between environmental parameters such as water temperature, dissolved oxygen levels, pH, and stocking density, and the biological state of the fish to achieve dynamic optimization throughout the growth cycle. Furthermore, for "quality prediction," by constructing an early biomarker panel based on molecular and metabolic data, we aim to transition from conventional post-management aquaculture to aquaculture that optimizes proactively based on predictions, enabling prediction of growth and quality before harvest.

This project is scheduled to be implemented from July to September 2026. During the target period, it will be promoted in five phases: data integration, AI model learning, optimization proposal, verification, and knowledge transfer.

Future Initiatives

Through this project, Keio Corporation and Nōka.AI will transform aquaculture sites, which have relied on "intuition and experience," with data and science. By stabilizing the production of Okutama yamame trout, we will ensure a continuous supply to restaurants and chefs seeking high-value ingredients. Furthermore, we will create an environment where chefs who value quality can choose Okutama yamame trout as an "always reliably available ingredient," further enhancing its value as a rare ingredient and contributing to the creation of a new food culture originating from the Keio line.

[Inquiries Regarding This Matter]

Keio Corporation My turn Operations Office: [email protected]

Keio Customer Center: 03-3325-6644 (9:00-18:00)

Nōka.AI Naka Honda: [email protected]

[Reference 1] About Keio Corporation

Comprising the Keio Line and Inokashira Line, its main business area is western Tokyo and northern Kanagawa Prefecture. As a life-related service provider, the group operates businesses in transportation, real estate, and hotels.

Representative: Satoshi Tsumura

Head Office Location: 1-9-1 Sekido, Tama City, Tokyo

Homepage: https://www.keio.co.jp/

Business Activities: Railway business, land and building leasing/sales, etc.

[Reference 2] Employee-Initiated Open Innovation Program

"Keio's Circular Cycle to Transform Restaurants into 'Enduringly Popular Establishments'"

Aiming to solve the problem of unstable supply due to abnormal weather and a shortage of successors for producers, despite the demand for ingredients with a story, we will stably provide restaurants with ingredients such as "Okutama yamame trout," "wasabi," and "basil" that possess added value such as environmental consideration and local production for local consumption, utilizing "aquaponics," a next-generation sustainable circular agriculture that combines hydroponics and aquaculture.

*Refer to the press release dated January 16, 2026, "Keio's Employee-Initiated Open Innovation Program 'My turn' Final Judging Held, Two Commercialization Projects Decided!!"

https://www.keio.co.jp/news/update/news_release/news_release2025/pdf/nr20260116_myturn.pdf

[Reference 3] About Nōka.AI, Inc.

A deep tech company that builds "virtual cell models" of organisms by fusing genome-scale metabolic models and multi-omics machine learning. Its main business areas include aquaculture, dairy farming, fermentation, livestock farming, and cell-based foods, promoting the advancement of biological understanding and industrial application through AI.

Japan Representative: Naka Honda

Head Office Location: California, USA

Homepage: https://www.nokabio.ai/

Business Activities: Development of a biological inference AI platform