Agata Co., Ltd. to Launch "RAG + AI Chatbot" Prioritizing Internal Data Protection, Accepting Applications from May 2026
Agata Co., Ltd., headquartered in Tomioka City, Gunma Prefecture, announced the launch of its new "RAG + AI Chatbot" service, which prioritizes the protection of confidential information through a closed-loop design that prevents external transmission of internal data. Applications will be accepted from May 2026. The service utilizes a self-owned data center powered by renewable energy, developed under a Ministry of the Environment project. It offers safe and sustainable business efficiency improvements for businesses of all sizes, from large enterprises to SMEs, at a low price of zero initial cost and 40,000 JPY (excl. tax) per month.
📋 Article Processing Timeline
- 📰 Published: May 13, 2026 at 00:00
- 🔍 Collected: May 12, 2026 at 15:32
- 🤖 AI Analyzed: May 12, 2026 at 16:21 (49 min after Collected)
Agata Co., Ltd., headquartered in Tomioka City, Gunma Prefecture (Representative Director: Yusuke Teshigawara), announced that it will start accepting applications from May 2026 for its new service, "RAG + AI Chatbot." This service thoroughly protects confidential information through a closed-loop design that does not transmit internal data externally. It utilizes a self-owned data center environment powered by renewable energy, developed under the Ministry of the Environment's "Zero Emission and Resilience Enhancement Promotion Project." At a low price of zero initial cost and 40,000 JPY (excl. tax) per month, it supports safe and sustainable business efficiency improvements for businesses of all sizes, from large enterprises to SMEs.
■ Background to Business Launch
In recent years, the business utilization of AI chatbots has rapidly advanced, but many systems transmit data via external clouds, causing many companies to hesitate in adopting them due to concerns about information leakage risks. Furthermore, societal interest in the power consumption and carbon emissions associated with AI operations is also growing.
Agata Co., Ltd. has already constructed its own renewable energy data center environment by utilizing the Ministry of the Environment's "Zero Emission and Resilience Enhancement Promotion Project." By adopting a closed-loop design that operates entirely on its own GPU servers, and blocking data routes to external clouds, it has minimized information leakage risks. The company aims to lower the barrier to adoption by offering zero initial cost, providing a foundation for companies of all sizes to utilize AI with peace of mind.
■ Four Strengths of This Service
1. Robust Security and Sustainable Infrastructure
This service is based on its own renewable energy data center environment. With a closed-loop design completed on its own GPU servers, data is not transmitted to external clouds. It strengthens corporate information management systems with a segregated operation that prioritizes confidential information protection.
2. "LLM Automatic Cleaning & Structuring" Function Reduces Implementation Workload
By concluding a non-disclosure agreement and entrusting internal documents such as Word and PDF files to Agata Co., Ltd., the service performs automatic cleaning and structuring using LLM for data with format deficiencies or unnecessary information. This data is then converted into a high-precision RAG environment (suitable for hybrid search + Reranker environment), enabling immediate practical use.
3. Continuous Automatic Updates for Latest LLMs Improving Performance
To keep up with the rapid evolution of AI technology, the service continuously updates to more accurate and faster latest LLMs after strict performance verification within the company. Current estimated design values maintain a response speed of 15-25 tokens/s and fast Time To First Token (TTFT), preventing stagnation in business workflows.
4. Transparent Pricing and Flexible Terms of Use
High-precision RAG functionality is available at a basic fee of 40,000 JPY (excl. tax) per month with zero initial cost. This price includes RAG processing for document data up to 100MB. One account supports one concurrent connection, but it can be shared within the company.
■ Pricing, Terms of Use, and Trial Service
・Basic Fee: 40,000 JPY / month (excl. tax)
* There is no token limit, but there is a usage count limit.
* This service adopts a model where initial investment is spread across monthly fees, and the pricing is based on the premise of continuous use for 2 years.
* Includes RAG processing for document data up to 100MB. Consultation is required for exceeding this amount.
・Usage Model: 1 account = 1 concurrent connection (shareable within the company)
・Contract Period: 2 years (Zero initial cost)
* By selecting a 2-year contract, it is offered at a special price of 40,000 JPY (excl. tax) per month.
・Early Termination Fee: Calculated based on the elapsed contract period (details in terms of service or contract).
・Trial Service: A 2-week free demo experience for an already created RAG chatbot is currently underway (applications being accepted).
As this service is in its initial offering phase, it will be provided to up to 500 accounts on a first-come, first-served basis. Please inquire early if you are interested. We also offer proposals for optimal usage tailored to your industry and application.
■ Representative's Comment
"In the business utilization of AI, securing data security and reducing environmental impact are by no means a trade-off. We have built our own managed GPU servers on a renewable energy infrastructure, eliminating the risk of confidential data leakage externally, while realizing sustainable and cost-effective AI utilization. We strongly hope that this service will be a solid first step for various companies to start their DX, and contribute to the realization of a society where more organizations can safely and sustainably enjoy the value of AI." (Representative Director Yusuke Teshigawara)
* Please note that while this service thoroughly implements multi-layered security measures, there are limits to any technical measures.
* This service is designed as a business assistance tool to support the rapid extraction and organization of internal information. The generated information should be used as reference material.
■ Background to Business Launch
In recent years, the business utilization of AI chatbots has rapidly advanced, but many systems transmit data via external clouds, causing many companies to hesitate in adopting them due to concerns about information leakage risks. Furthermore, societal interest in the power consumption and carbon emissions associated with AI operations is also growing.
Agata Co., Ltd. has already constructed its own renewable energy data center environment by utilizing the Ministry of the Environment's "Zero Emission and Resilience Enhancement Promotion Project." By adopting a closed-loop design that operates entirely on its own GPU servers, and blocking data routes to external clouds, it has minimized information leakage risks. The company aims to lower the barrier to adoption by offering zero initial cost, providing a foundation for companies of all sizes to utilize AI with peace of mind.
■ Four Strengths of This Service
1. Robust Security and Sustainable Infrastructure
This service is based on its own renewable energy data center environment. With a closed-loop design completed on its own GPU servers, data is not transmitted to external clouds. It strengthens corporate information management systems with a segregated operation that prioritizes confidential information protection.
2. "LLM Automatic Cleaning & Structuring" Function Reduces Implementation Workload
By concluding a non-disclosure agreement and entrusting internal documents such as Word and PDF files to Agata Co., Ltd., the service performs automatic cleaning and structuring using LLM for data with format deficiencies or unnecessary information. This data is then converted into a high-precision RAG environment (suitable for hybrid search + Reranker environment), enabling immediate practical use.
3. Continuous Automatic Updates for Latest LLMs Improving Performance
To keep up with the rapid evolution of AI technology, the service continuously updates to more accurate and faster latest LLMs after strict performance verification within the company. Current estimated design values maintain a response speed of 15-25 tokens/s and fast Time To First Token (TTFT), preventing stagnation in business workflows.
4. Transparent Pricing and Flexible Terms of Use
High-precision RAG functionality is available at a basic fee of 40,000 JPY (excl. tax) per month with zero initial cost. This price includes RAG processing for document data up to 100MB. One account supports one concurrent connection, but it can be shared within the company.
■ Pricing, Terms of Use, and Trial Service
・Basic Fee: 40,000 JPY / month (excl. tax)
* There is no token limit, but there is a usage count limit.
* This service adopts a model where initial investment is spread across monthly fees, and the pricing is based on the premise of continuous use for 2 years.
* Includes RAG processing for document data up to 100MB. Consultation is required for exceeding this amount.
・Usage Model: 1 account = 1 concurrent connection (shareable within the company)
・Contract Period: 2 years (Zero initial cost)
* By selecting a 2-year contract, it is offered at a special price of 40,000 JPY (excl. tax) per month.
・Early Termination Fee: Calculated based on the elapsed contract period (details in terms of service or contract).
・Trial Service: A 2-week free demo experience for an already created RAG chatbot is currently underway (applications being accepted).
As this service is in its initial offering phase, it will be provided to up to 500 accounts on a first-come, first-served basis. Please inquire early if you are interested. We also offer proposals for optimal usage tailored to your industry and application.
■ Representative's Comment
"In the business utilization of AI, securing data security and reducing environmental impact are by no means a trade-off. We have built our own managed GPU servers on a renewable energy infrastructure, eliminating the risk of confidential data leakage externally, while realizing sustainable and cost-effective AI utilization. We strongly hope that this service will be a solid first step for various companies to start their DX, and contribute to the realization of a society where more organizations can safely and sustainably enjoy the value of AI." (Representative Director Yusuke Teshigawara)
* Please note that while this service thoroughly implements multi-layered security measures, there are limits to any technical measures.
* This service is designed as a business assistance tool to support the rapid extraction and organization of internal information. The generated information should be used as reference material.