AI Development in Kyoto

Tomohiro Iida · Published June 28, 2026 · Updated July 11, 2026

If you're looking to talk to someone about AI development or generative AI adoption in Kyoto, a company that both implements AI in real operations and runs its own AI and data R&D is one option. Netsujo, based in Kyoto, tells you honestly what is and isn't possible, and starts from a small PoC to confirm the effect before moving forward.

The arrival of generative AI has brought more requests to apply AI to a company's own operations. This article covers what to check before commissioning AI development or generative AI adoption work in Kyoto, what Netsujo works on, and how we proceed.

This article is aimed mainly at AI adoption and applying generative AI. If you're looking for contract development (choosing a vendor) for a business system or app, see our article on system development and contract development in Kyoto.

01. What to check before commissioning AI development

Checking three things before commissioning AI development — whether the issue is one AI can actually solve, verifying the effect with a PoC, and designing how it fits into existing operations and systems — helps avoid a mismatch after adoption.

Judge whether AI is the right tool for the problem

AI is not a universal solution. For work with clear, fixed rules, conventional automation can be a better fit. We start by working out what specifically requires AI, and scoping where it applies.

Verify the effect with a PoC before committing

Before full-scale development, we check accuracy, cost, and operational realism with a small PoC. Generative AI's effectiveness depends on the input data and the workflow, so verifying first makes the investment decision safer.

Design how it fits into existing operations and systems

Beyond AI accuracy, how it fits into the existing workflow and systems is what determines whether it actually sticks. We consider the data pipeline, permissions, and operating structure from the design stage.

02. Netsujo's work on AI

Netsujo works on AI across three areas: implementing AI in operations, our own R&D, and supporting information delivery for the AI-search era (SIGNAL).

Implementing AI in operations and supporting generative AI adoption

We work on operational efficiency using generative AI (LLMs), building data analytics infrastructure, adding AI features to web services, and RAG and chatbot development. Details are on our AI social implementation service page.

Our own R&D

In decentralized identity (DID), we continue joint research and development with Proof of Your Life, Inc. Working on our own R&D gives us a grounded view for helping clients judge what is realistic.

Supporting information delivery for the AI-search era (SIGNAL)

We also offer Netsujo SIGNAL, which supports information design (AIO/SEO) built to be cited by AI search such as AI Overviews, ChatGPT, and Perplexity. We support both AI adoption and information delivery.

03. How we proceed with AI development and adoption

We start with a consultation (working out what you want to do and the issue), verify the effect with a PoC, and move step by step into implementation and operation. Rather than building something large from the start, we confirm things in small steps and make the investment decision as we go.

We aim to send a first reply within three business days of an inquiry. We also work with you at the stage where requirements are not yet settled.

Frequently asked questions

Can I talk to someone in Kyoto about AI development or generative AI adoption?

Yes. Kyoto-based Netsujo accepts consultations on implementing AI in operations or adopting generative AI, starting from the concept stage. We also work with you at the stage where there is something you want to do but you are not sure how to proceed.

How can generative AI be used in day-to-day operations?

It can help with summarizing and classifying documents, assisting with inquiry response, searching internal knowledge (RAG), and supporting data analysis. That said, leaving judgments that require strict accuracy entirely to AI is not appropriate. We work out what fits and what does not up front.

What scale of project can I start with for AI development?

You can start from a small PoC. We recommend confirming the effect within a limited scope first, then deciding whether to move to full-scale development based on the result.

Will you also tell me what AI can't do?

Yes. To avoid overpromising, we honestly share where AI falls short, and the limits of accuracy and operation. We share what is and isn't possible up front so you don't end up finding a gap between expectation and result after adoption.