————.——.—— ——:—— — TOKYO
JA KO EN
Service

Not consulting alone,
not implementation alone. Strategy meets execution. Visualize, then ship.

01 MAIN — Service

AI Automation &
Operations.

Pain — what we keep seeing
  • Are you still using FAX?
  • More tools, but no less work?
  • Are people manually wiring tools together?
Answer — AISolve's response

That inefficiency — AI can solve it.

RPAAI-OCRWorkflowSlack/TeamsClaude/GPTMCP
Related cases Case 01 — Process automation →Case 02 — AI organization (in-house proof) →Case 03 — AI enablement training →
02 Service

Tech Advisor /
CDO-in-residence.

Pain — what we keep seeing
  • Want to drive DX, but no one in-house to discuss it with.
  • IT conversations don't connect to executive decisions.
  • Can't tell whether vendor proposals are reasonable.
Answer — AISolve's response

An external CTO/CDO who speaks management's language, with 20 years on the floor.

AdvisoryArchitecture ReviewTech Due DiligenceDX Strategy
Related cases Case 02 — AI organization (in-house proof) →
03 Service

App & Web
Development.

Pain — what we keep seeing
  • Off-the-shelf SaaS doesn't fit our operation.
  • We want a custom tool, but big vendors quote too high.
  • Want AI but don't know where to embed it.
Answer — AISolve's response

We build your AI-embedded tools, small and fast.

React NativeTypeScriptCloudflareFirebaseClaude API
Related cases AI changed what counts as normal at work. →
04 Service

Korea → Japan
Market Entry.

Pain — what we keep seeing
  • Want to enter Japan but don't know where to start.
  • Japanese business norms turned out more complex than expected.
  • Local partner is in place, but is it really the right fit?
Answer — AISolve's response

From market sizing to IT foundation — we ride along.

Market EntryJP-KR TranslationLocal Partner SourcingIT Setup
Works

Automate before you hire.

See work as lines, not dots. A set of lines becomes an organization. (dots → lines → organization)

Case 01 Process automation
End-to-end AI workflow for customer support

Customer support, end to end with AI.

Gmail, kintone, an FAQ site — are your people acting as the middleware between disconnected dots? We treat work as a line, and connect it from inbound message to reply.

Before After
Staff key inbound emails into the CRM by hand AI parses inbound email and registers it in the CRM automatically
Staff search the FAQ by eye and draft replies from scratch AI matches against the FAQ and drafts the reply
Every hand-off between tools is manual (copy and paste) People only do the final check and send
Effect

Email handling effort down roughly 70% (email channel basis)

AIn8nClaudeCRM integrationEmail processing
Related service AI Automation & Operations →
Case 02 AI organization (in-house proof)
Org chart: AI departments audited by a QA AI, with a human giving final approval

An organization isn't headcount.

The functions a company needs can be built, not hired.

Marketing, accounting, engineering. Does all of that really have to be done by hand? At AISolve we build the functions a company needs as AI "departments," and prove them on ourselves every day.

Department What the AI actually does (in production)
Marketing Plans and writes articles weekly. People only approve
Consulting Tracks the deal pipeline, alerts on follow-up deadlines, drafts proposals
Administration AI classifies invoice emails, then after human approval the system posts them to our accounting software (freee). Monthly close reports and 30/60/90-day cash-flow forecasts
Engineering Product development and weekly progress reports
Quality assurance A separate AI audits the first AI's output before it reaches a human for approval
Effect

Back-office run with no dedicated staff / executive briefings and cash-flow forecasting automated

ClaudeGASfreeeMulti-agentApproval gate design
Related service AI Automation & Operations →Tech Advisor / CDO-in-residence →
Design principle

AI decides, the system executes, a person signs off. "But doesn't AI make mistakes?" — that is exactly why we place an auditing AI in the loop and require a human for final approval.

Case 03 AI enablement training
AI enablement training where every participant is working hands-on

Leave no bystanders.

Training you only watch ends in a shrug, and nothing changes the next day. So everyone puts their hands on the keyboard — in a form you can use tomorrow, even with no IT staff on site.

Case (HR consulting firm, custom program)

1) Lecture (the history of AI and how to approach it) → 2) Hands-on demo (app building, music generation) → 3) Practice (teams of two to three complete a proposal for a fictional client, same day).

For the practice session we pre-loaded a fictional company profile and the client's own consulting method, then added interview notes on the day. Each team worked through the proposal with AI and produced the proposal document itself.

Effect

What we hear afterwards — "Our people changed from that day. Everyone wants to try things, everyone wants to change."

Additional track record

Company-wide training for a company listed on the Tokyo Stock Exchange TOKYO PRO Market (TPM) among others, with a consistent design from the executive team down to the front line. We deliver the same design at the scale of a smaller business.

Generative AI trainingWorkshop designCustom materialsPrompt design
Related service AI Automation & Operations →

* Some of these cases were planned and delivered by our founder while at their previous employer.

Proof

AI changed what counts as normal at work.

The old assumption Now (proven in-house at AISolve)
A website is built once, then costs you updates and upkeep AI writes an article every week, published dynamically at near-zero running cost (ai-solve.net) See it — Insights on this site →
Building an app takes a team, a big budget and a long timeline One person, small and fast, all the way to store release (Kumitate Hangul — available on Android; iOS release date to be announced) See it — Kumitate Hangul →
Kumitate Hangul app screen
AISolve's first product

Kumitate Hangul

It started while teaching my son Hangul: there aren't many apps that let you learn it structurally. So I decided to build one myself — Kumitate Hangul.

A Hangul-learning app built on React Native / TypeScript / Firebase, designed so the structure of Hangul can be learned in a short time and without strain.
Even absolute beginners get an intuitive grasp of the character structure. AISolve's first product, released on Google Play on July 12, 2026. An iOS version is planned (timing to be announced).

React NativeTypeScriptFirebaseClaude APIAndroid live / iOS planned

"It'll be expensive." "It'll take forever." Those assumptions may pre-date AI. Start with 30 minutes, looking at your invisible inefficiency together.

Book a free consultation→
00 · 30-Second Diagnose
⏱ avg. 30 sec

What's actually
happening atyour company right now?

Let's start by
looking togetherat where you actually stand.

Book a free consult →