We stay with you until AI becomes
part of how your team works.
AI producers, product managers and engineers with experience in new business development and AI system development join your team on site as FDE（Forward Deployed Engineer）, and support you from defining the challenges to implementation and adoption. We cover the introduction of generative AI, business optimization, AI product development and proofs of concept (PoC). Our goal is not just to try AI, but to turn it into something your company keeps using.
-
Strength 01
Listen, break down, and put it into words
Editors with interviewing and editing experience, led by our CEO, who served as deputy editor of an owned media outlet at an operating company, draw out even the issues that frontline staff have not yet voiced. We break down and organize the issues with project management methods, and put into words what AI is meant to solve before we proceed.
-
Strength 02
An executive and engineer who works hands-on as an FDE
The people leading AI adoption are members who founded and run their own companies while also developing as AI engineers. As FDEs (Forward Deployed Engineers), they work inside a company’s team and see it through together, from proposal to design, development, and in-house training.
-
Strength 03
Bring AI we built and used heavily ourselves
At Cognify Lab., in which our members take part, we develop and operate 16 AI products. They include structuring internal knowledge (Nolvia), creating proposals and quotes (Axion), lodging operations (NEWO SWITCH), and corporate talent development (STARTOUT). We bring what we learned by running everything ourselves, from prototype to operation, about where to use AI and where people should check, into our support.
Not sure where to start with AI?
Let’s figure that out together.
“We’d like to try AI, but we don’t know what to use it for.” Please come to us at that stage. We listen without technical jargon, find a way of using AI that fits your company’s work together with you, and help you until your team can use it on its own. Consultations are free.
Clients
Clients we have worked with
Worries
Do any of these sound familiar?
- We were told to “use AI,” but we don’t know what to use it for
- We tried ChatGPT, but we haven’t been able to use it at work
- We’re not sure whether it is safe to enter internal information, and we worry about information leaks
- No one in our company knows much about AI
- We don’t know how much it will cost
If even one applies to you, start with a free consultation.
What we do
How Number X can help
- 01
We find the right uses for AI together
We learn how your work runs today and look for tasks that get easier when AI takes them on. We will also tell you plainly which work is better done without AI.
- 02
We get everything ready to use
We choose AI tools that fit your purpose, and handle setup and the creation of internal rules. If information leaks are a concern, we decide on safe ways of using AI together.
- 03
We stay with you until it sticks
We hold briefings for employees, prepare how-to materials, and answer questions after introduction. We won’t let it end up as “introduced but unused.”
Examples
Examples of what gets easier
-
Drafting text
NowWriting every email and announcement from scratch
With AIAI writes the draft, and people only review and edit it
-
Replying to inquiries
NowLooking things up every time to answer similar questions
With AIAI proposes an answer based on internal documents
-
Meeting summaries
NowWriting the minutes after a meeting takes a long time
With AIAI summarizes the key points from the recording, and a person checks them
-
Finding documents
NowNot knowing where the documents we need are
With AIAsk AI, and it tells you which documents apply and where they are
These are just a few examples. We will work out how to use AI together, based on your work.
Flow
How consultations work
- STEP 1InquiryUse the form below to send us what is troubling you, even in a single line.
- STEP 2Free consultationWe listen to how your work runs and what troubles you, without technical jargon. Consultations are free.
- STEP 3Proposal and startWe propose what to work on, how to proceed, and an estimate. We begin only after you are satisfied.
Subsidies and grants may be available
National and local government subsidies and grants may be available for introducing AI tools and training employees. We help you with everything from identifying which programs you may be able to use to preparing the contents of your business plan.
FAQ
Frequently asked questions
Is consultation free?
Yes, consultations are free. After hearing from you, we propose what to work on and a quote.
Is it all right if we do not know much about AI?
Yes. We explain without technical jargon and help until your staff can use it.
Can we consult even if nothing has been decided?
Yes. We welcome consultations from the stage of "we do not know what to use it for."
Can a small company ask for help?
Yes. We also take consultations from small companies and startups.
We are worried about information leaks.
We help from creating internal rules on what information may be entered and which tools may be used. Together we decide a safe way of using AI that fits your company’s environment.
How much does it cost?
It depends on the scope of work, so we prepare a quote after hearing from you. We also advise on whether subsidies or grants can be used.
Service
Main services
-
01Organize
We cannot decide what to use AI for
AI use issue assessment and use case design
We listen to frontline work and challenges, and sort out tasks AI can take over, judgments AI can assist, and areas where it is better not to use AI. Based on expected effect and effort, we decide the first use case to tackle.
- Interviews on work and challenges
- Listing and prioritizing use cases
- Plan for the order of work
-
02Visualize
Work depends on individuals, and we do not know where to start
Business process visualization and optimization
We write out the different procedures each person follows and map the flow of work. We separate tasks for AI from those for people to decide, and review the procedures themselves.
- Creating workflows
- Identifying manual and duplicated work
- Designing the scope handed to AI
-
03Adoption
We introduced a tool, but nobody in the company uses it
Generative AI tool selection, adoption, and in-house rollout support
We select generative AI tools that fit your purpose and internal environment (security, existing systems), and handle setup and adoption. With briefing sessions and usage guides, we get the tools used on the front line.
- Tool comparison and selection
- Initial setup and adoption
- Internal briefings and usage guides
-
04Development
Off-the-shelf tools do not fit our business
Planning, design, and development of AI products and business tools
We plan AI business tools tailored to your work and AI products embedded in your own services, and take charge from screen design through development. With AI-driven development using AI coding tools, we shorten the time from prototype to release. We also support introducing AI-driven development to development teams.
- Planning and organizing requirements
- UI/UX design
- Development and release through AI-driven development
- Introducing AI-driven development to development teams
-
05Validation
We want to confirm the effect before full adoption
Design and execution of proof of concept (PoC)
We build a prototype on a small scale and check the effects and issues. We decide the items to validate and the criteria for judgment beforehand, and gather what you need to decide whether to move to full adoption.
- Validation design (scope, indicators, criteria)
- Prototyping and execution
- Reporting results and the next plan
-
06Operations
We worry about information leaks and misuse
Usage rules and operating structure design
We create usage rules covering what information may be entered, how to check output, and which tools may be used. We decide owners and review timing, and set up a structure you can keep running.
- Creating usage guidelines
- Designing the operating structure and roles
- Regular reviews
AI agents we combine
Instead of leaving everything to a single AI, we split the work into components by role (agents) and combine them. We also decide at the start where people will check the results.
-
InputGather
- ResearchGather the information you need from internal and external materials and the web
- CollectRegularly crawl set sites and data to bring in new information
-
ProcessThink
- DecomposeSplit work and requests into tasks that AI and people can each take on
- Extract and structurePull needed items out of documents and notes and arrange them as tables and data
- DraftCreate drafts of text and documents following set formats and style rules
- JudgeSort and pass or fail with rules or an LLM, and record the reasons
-
CheckVerify
- Verify and fact-checkCompare against sources and original text to find errors in facts and numbers
- Human approvalA person checks before irreversible actions such as publishing, sending, or going live
-
OutputConnect and monitor
- API integrationConnect to internal systems and SaaS and automate data handoffs
- Monitor and notifyWatch for changes and anomalies, and notify the person in charge only when needed
Examples of how agents are combined
-
Creating a press release
- Research (past releases, style rules)
- Writing and drafting (body text, headline ideas)
- Verification and fact-checking (style, facts)
- Human approval
-
Questions to internal knowledge
- Collection (regulations, manuals)
- Extraction and structuring
- Answer (with sources)
- Monitoring and alerts (record questions that could not be answered)
-
How the brand appears in AI search
- Research (ask AI search)
- Extraction and structuring (mentions, cited sources)
- Judgment (changes since last time)
- Monitoring and alerts
- Improvement proposals
-
A Number X example: analyzing and improving search data
- Analysis (read Search Console and GA4 with Codex)
- Breakdown (choose what to fix)
- Implementation (Claude Code)
- Verification (check with Codex)
- Record (use for the next comparison)
-
A Number X example: redesigning this site
- Breakdown (turn requests into numbered tasks)
- Research and implementation
- Verification (check display and accessibility by machine)
- Human approval
- Release to production
Flow
How we work (an example)
- STEP 1 Clarify issues We listen to frontline work and challenges, and decide the use cases to tackle with AI.
- STEP 2 Visualize work We map the flow of work and separate what AI handles from what people judge.
- STEP 3 Proof of concept (PoC) We try on a small scale and check the effects and issues.
- STEP 4 Adopt and develop We introduce generative AI tools and, as needed, develop business tools and AI products.
- STEP 5 Rollout and operations We prepare briefings and usage rules so your company can keep using it.
Subsidy
[Coming soon] AI Implementation Subsidies and Grants
National and local government subsidies and grants may be available for introducing AI tools and training employees. Along with AI implementation support, we also help you prepare your applications. We stay with you from identifying the programs you can use to preparing the implementation plan, the estimate and the contents of the business plan.
- 01Find the programs you can use
- 02Prepare the implementation plan and estimate
- 03Develop the contents of the business plan
- 04Implementation and reporting after approval
-
Small and Medium Enterprise Agency; Organization for Small & Medium Enterprises and Regional Innovation, JAPAN
Digitalization and AI Adoption Subsidy (Digital-ka / AI Donyu Hojokin; formerly the IT Introduction Subsidy, IT Donyu Hojokin)
Small and medium-sized enterprises and small businesses introducing IT tools, including AI, to make operations more efficient
Applications require IT tools registered with the secretariat and must go through an IT adoption support provider.
Official site（Digitalization and AI Adoption Subsidy (Digital-ka / AI Donyu Hojokin; formerly the IT Introduction Subsidy, IT Donyu Hojokin), external site) -
Ministry of Health, Labour and Welfare
Human Resources Development Support Subsidy (Jinzai Kaihatsu Shien Joseikin)
Training employees (reskilling), such as how to use generative AI
Courses include the Business Expansion Reskilling Support Course (Jigyo Tenkai-to Reskilling Shien Course). A plan must be submitted before the training.
Official site（Human Resources Development Support Subsidy (Jinzai Kaihatsu Shien Joseikin), external site) -
Prefectures, cities, and municipalities
Local government subsidies and grants
Getting support from local programs for introducing DX or AI, using experts, and more
Programs differ depending on where your company is located. You can search for them on a national support information site.
Official site（Local government subsidies and grants, external site)
Subsidy rates, upper limits and deadlines vary by fiscal year and application round. Please check the latest application guidelines and the guidance from the program secretariat and the Labor Bureau. Preparing and submitting application documents on someone else’s behalf is work reserved for professionals such as social insurance and labor consultants (shakai hoken romushi) and administrative scriveners (gyosei shoshi), so Number X does not do it. If you need that service, please consult a professional.
Use Case
Use cases
What kinds of companies can use AI, for what kinds of work, and how. Here are some representative examples.
-
Make routine work more efficient
Real estate company running vacation rentals and lodging
One place for reservations, guest messages, and cleaning
In vacation rental and lodging operations, booking sites, guest messages, cleaning notices, and task management tend to be scattered across different places. NEWO SWITCH from Cognify Lab. is a web app that brings them onto one screen. It pulls properties, reservations, and messages from the API of the reservation management service Beds24, and AI drafts the replies.
- 01
Handle messages in one place
Guest messages are lined up in threads and can be filtered by property. Translating incoming text, drafting replies with AI, editing, and scheduling the send all happen on the same screen. Sends are scheduled for five minutes later and can be canceled.
- 02
Expand automatic replies step by step
AI refers to each property’s rules to judge whether it may reply. It tries automatic replies on low-risk inquiries first, and routes high-risk or human-handled matters to a pending-review queue or tasks. The scope of automation is switched with an overall switch and per-property settings.
- 03
Connect cleaning and tasks to reservations
Unanswered inquiries, equipment problems, and cleaning shortfalls become tasks whose priority, owner, deadline, and status can be tracked. Cleaning schedules are created from reservation data, and progress is checked in monthly, weekly, and daily calendars.
NEWO SWITCH is an MVP (prototype) under development at Cognify Lab. This is an example assuming its features are introduced at a company with the same challenges.
- 01
-
Make routine work more efficient
PR department of a manufacturer
A system for drafting press releases with AI
Based on past releases and style rules, AI creates draft text and headline ideas, and a person reviews and finalizes them. We set up this flow.
-
Improve how the brand appears
Brand division of a consumer goods maker
Monitoring and improving how your brand appears in AI search
We regularly check how the brand is treated in AI search answers and which sources are cited, and improve the website and communication accordingly.
-
Make resident services more efficient
Resident service counter of a city or ward
An FAQ that answers resident inquiries with sources
We organize procedure guides, ordinances, and guidelines into knowledge AI can work with, and create draft answers with sources shown. Staff check them before giving guidance.
-
Deliver information overseas
Tourism and inbound travel company
Making a WordPress site multilingual
AI translates published articles and pages, and we set up per-language URLs and search settings (hreflang, sitemaps).
-
Standardize proposal work
Consulting firm or agency
A system that creates proposals and quotes from client interviews
From meeting materials and interview content, AI creates drafts of AX/DX initiatives, quotes, and proposals.
-
Build AI into the product
Startup working on a new business
Rapid prototype (MVP) development and validation with AI
We build a prototype in a short time and, through a proof of concept, check how users respond before moving to full development.
All of these are hypothetical examples, not actual implementation cases. The approach and the effects vary depending on the work and the environment.
Works & Lab
AI products built into business operations
Structuring internal knowledge, creating proposals and estimates, managing accommodation operations, organizing the profitability of e-commerce, and developing corporate talent. We turn the challenges found on the front lines of business into tools and products with AI built in. These are 16 projects developed and operated at Cognify Lab., in which our members take part. Each card opens its detail page (external site).
Each project is owned by the individual, company or organization that created it.
Products
Our own services built with AI
These are services developed and operated by the Number X group through AI-driven development with AI coding tools. From the systems that collect and organize public information to the search and comparison screens, we build them together with AI.
- genai-db.com GenAI Tool DB A database to search and compare AI tools and AI SaaS around the world by "what I want to do," features, industry, pricing, Japanese-language support, and more. How AI is usedThe pipeline runs from data collection through validation to publication. Every change by AI agents is recorded and can be rolled back, and important items are applied only after a person approves them. Visit site：GenAI Tool DB (external site)
- realshop-checker.com Realshop Checker A service that checks, from public information such as official websites and store listings, whether a restaurant found on Uber Eats or Demae-can has a physical location. How AI is usedIt automatically gathers multiple pieces of public information and judges with an LLM (Claude). It always records the reason for the judgment, and switches to rule-based judgment when the LLM is unavailable. It re-checks automatically once a day. Visit site：Realshop Checker (external site)
- leathercare-db.com LEATHER CARE DB A database that links leather goods such as leather shoes and bags to the care products usable on their materials, based on manufacturers’ public information. How AI is usedThe LLM is used only to extract information from manufacturers’ public information, and compatibility is decided by rules. We verify by machine that extracted content matches the quoted source text, and it is updated automatically by scheduled jobs. Visit site：LEATHER CARE DB (external site)
- luxury-hotel-db.com STAY LOG A database of lodging experiences at luxury hotels and ryokan, including introductions, stay reports, and room tours. How AI is usedA collection agent finds and updates hotels, stay reports, and videos. We do not store the text of external articles or create replacement text with AI summaries, in consideration of copyright. Visit site：STAY LOG (external site)
Operated by: Barrel Connect, Inc. (a new company of Number X)
Member
AI Implementation Support members
FDE / AI Engineer / AI DesignerShota Kamiyama
Began his career in 2007 in web design and software engineering. Building on digital creative work, web development, and UI/UX design, he gained practical expertise spanning planning, production, and operations. He then moved to the United States, and after returning to Japan he launched an entrepreneur development program and worked in new business development, entrepreneurship education, and human resource development. He also provides business support across many fields, including marketing strategy planning, brand design, and digital product development support. In recent years, he has focused on the potential for putting artificial intelligence technologies, including generative AI, to practical use in society, and for about two years he has intensively researched and practiced AI utilization and AI system development. He currently provides hands-on AI implementation support, business optimization support, and AI product development support to companies and business operators. He also founded QUON Inc., which, in addition to AI implementation support and web production and development businesses, promotes a community-building and village-building project to nurture creators and entrepreneurs, based in Japan and in Bali, Indonesia.
Assets:#BusinessDevelopment #AI
CEO / AI Producer / FDE / Project ManagerTsubasa Yasuda
Number X, Inc. CEO / DeeTeller, Inc. CCO. As a web producer, project manager, and growth hacker at a design consulting firm (now part of the Monstarlab Group), worked on owned media, websites, and content production for companies listed on the First Section of the Tokyo Stock Exchange. From 2018, at SoftBank Corp.’s Public Relations Division, worked as deputy editor-in-chief on project management, growth hacking, and editing for owned media. Alongside that work, founded Number X, Inc. (Number X) in March 2021. Became Director and CCO of DeeTeller, Inc. in February 2022. From 2022 to 2023, served as Director and CMO of YourTrade, Inc., a startup running a cross-border business.
Assets:#PR #OwnedMedia #GrowthHacking #Producing #ProjectManagement #WebDirection #IA #UI #UX #WebAnalytics #Editing #B2BMarketing
Technical Director / Forward Deployed EngineerKoya Ikeda
In 1995, at the university he was attending, he encountered the web browser “NCSA Mosaic” and was struck by it. In 1997, he moved to the United States to study IT. In 2003, he graduated from the University of California, San Diego, earning a degree in computer science from the Jacobs School of Engineering. After returning to Japan, he worked in sales and marketing at VA Linux Systems Japan K.K., the Japanese subsidiary of a U.S. technology venture. In 2007, he became an independent freelance engineer and joined various new business projects, mainly in Tokyo, as the technical lead. In 2012, he founded Digirou Inc. (formerly ClearTouch Inc.) in Utsunomiya, Tochigi.
Assets:#AI implementation support #AI-driven development #FDE #Technical direction
Product Manager / Project ManagerKeisuke Nagasoe
Has handled direction and project management on many large, medium, and small projects. With varied experience as a consultant, director, engineer, designer, and in sales and marketing, he has driven many web development and operations projects. Most recently, as a product manager, he has worked on direction, project management for large projects, and business improvement consulting. He has extensive experience as a project leader, including overseeing partners from multiple companies.
Assets:#BusinessImprovementConsulting #ProductManagement #ProjectManagement #Direction #WebDirection #IA
Contact
For a free consultation about AI implementation support, please use this form. “Project inquiry” is already selected as the type of inquiry.