Two Deployment Options
Implementation track record
User retention rate after onboarding
*As of Mar 2026, Retention rate among companies that have implemented the system, including those with a PoC (Proof of Concept).
Features / Functions
Integrating multiple cutting-edge LLMs (Large Language Model) such as ChatGPT, Gemini, and Claude, we provide a secure environment where all employees can safely use conversational AI without their input data being used for AI training.
Securely link and integrate client companies’ unique data. Using highly accurate RAG technology (Retrieval-Augmented Generation), we build a unique search system that derives precise, actionable answers from text and images (Multimodal).
Analysis of usage, costs, accuracy evaluation, and other analyses required by implementation managers are possible.
We provide AI Agent functionality that autonomously executes specific business processes. We can also develop and extend custom agents optimized for your company’s complex individual tasks.
Know Narrator's Three Strengths
"Highest Level of Security" to Protect Confidential Information
We provide a secure AI environment that can be confidently implemented even in companies with extremely strict security policies, such as financial institutions and manufacturing companies.
- Construction of a company-specific “closed space” on the cloud
- Adoption of “Azure OpenAl Service” where input data is not used for AI training*
"Unrivaled RAG Accuracy Optimized for Data Characteristics"
Even with default settings, it achieves highly accurate answers while suppressing hallucination (plausible lies). Furthermore, it applies optimal processing and tuning to company-specific data, elevating it to a “next level of accuracy” that sets it apart from general-purpose tools.
- Information retrieval and answer generation algorithm supervised by a “Kaggle Master,” a mark of world-class AI engineers.
- Implementation know-how cultivated through hundreds of generative AI projects.
- Advanced customization of chunks and models according to data characteristics.
Superior Support that Balances "Thoroughness" and "Minimalism"
Standard features that meet enterprise requirements enable smooth implementation without support. However, each company’s prerequisites (data, systems, use cases) are vastly different. For “accuracy improvement x system integration,” which is difficult to solve with general-purpose SaaS, we provide “superior support” based on our solid track record cultivated over many years in enterprise SI, offering thorough yet minimalist approach to deeply adapt to the client company’s business environment.
- Enterprise-level standard features that can be implemented and utilized even without support
- Deep business understanding gained from supporting leading Japanese companies for 50 years
- System scalability to meet customer requirements (API integration and full-scratch development experience)
Service Comparison
How Know Narrator compares against general global services and domestic SaaS alternatives.
| Features / Capabilities |
|
General global services | Similar domestic SaaS services |
|---|---|---|---|
| Security |
◎
Robust enterprise-grade, private network capable.
|
◎
Robust enterprise standards
|
△
Vendor dependent
|
| Supported AI Model |
◎
Supports multiple protocols + latest models (ChatGPT, Gemini, Claude, etc.)
|
△
Principle: Single
|
〇
Multiple support. There is a time lag in implementing the latest version.
|
| Search and response accuracy |
◎
Standard high precision + advanced tuning possible. Multimodal RAG compatible.
|
△
Precision tuning is limited.
|
〇
Basic accuracy
|
| Internal data integration |
◎
Integration with diverse data sources is possible.
|
△
Only specific data integrations
|
〇
Some restrictions apply.
|
| Business system integration |
◎
Offers diverse APIs and allows for flexible integration into business processes.
|
△
Only within the company's ecosystem
|
〇
Basic API provision
|
| Individual customization |
◎
Flexible SI development support
|
△
This can be handled by external vendors, etc.
|
〇
Handled within the default range
|
| Implementation speed |
〇
From requirements definition to service launch within one month
|
◎
Immediate use with additional license
|
◎
Available immediately or within 1 month
|
*This comparison table is based on our own research and evaluation using publicly available information. The specifications and services offered may change without notice.
Frequently Asked Questions
Q. I'm worried about using AI on the cloud because it handles confidential company data.
Q. Is there a risk that the data I input will be used for AI training?
Q. Our internal documents have inconsistent formats and are disorganized. Do we need to organize them beforehand?
A. Our experienced data scientists will provide comprehensive support, from data organization before implementation to actual operation and adoption, tailored to your needs.
Please rest assured.
Q. Can we integrate AI into our existing internal portals and chat tools?
A. Yes, it is possible. Through Know Narrator’s diverse APIs, we can seamlessly integrate it with your existing business systems and external services.
Because you can directly invoke AI from your daily business systems and applications, you can improve the efficiency of your daily operations.
Furthermore, Dentsu Soken can also handle the development for this integration.
Q. How do I choose a generation AI service?
Q. How much will it cost?
A. The cost depends on the features and configuration you use, but it starts from several hundred thousand yen per month. (Initial setup fees apply separately.)
We will propose the optimal pricing plan to suit your usage.
Q. After implementation, will the IT department's burden increase due to internal inquiries and AI accuracy adjustments?
Q. I'm interested, but I don't know how to proceed. Where should I start?
A. Dentsu Soken provides support from the initial stages of consideration, including identifying challenges and clarifying requirements for implementation. Please feel free to contact us for a consultation.
We also offer a “Generative AI Promotion Guide” for companies struggling with the use of generative AI. Please feel free to use it for your internal discussions.
*The handling of input data complies with Microsoft’s latest terms of service and privacy policy.
*Due to the nature of this service, which utilizes generation AI technology, we cannot guarantee the completeness of the output responses. For actual business applications, please perform final verification by a human as needed.
*”ChatGPT” is a trademark or registered trademark of OpenAI OpCo, LLC. “Claude” is a trademark or registered trademark of Anthropic, PBC. “Microsoft,” “Azure,” and “Azure OpenAI Service” are registered trademarks or trademarks of Microsoft Corporation in the United States and other countries. “Gemini,” “Vertex AI,” “Kaggle,” and “Kaggle Master” are trademarks or registered trademarks of Google LLC. Other company names and product/service names mentioned are registered trademarks or trademarks of their respective companies.
