Kan Ninomiya
JA

Profile

Kan is a 20-year-old software engineer and product builder in Hiroshima, Japan. He builds AI agents, Japanese language and speech systems, and web and mobile products, then runs what he ships.

Basic information

Name
Kan Ninomiya
二宮 貫
ニノミヤ カン
Age
20 years old
As of 2026-07-10
Roles
Representative Director and President / COO at FP16 Inc.
Software Engineer
Product Builder
Base
Hiroshima, Japan
Personality
ENFP

Biography

Kan Ninomiya is a 20-year-old Hiroshima-based software engineer and product builder working on AI agents, web and mobile products, and Japanese language and speech systems. As president and COO of FP16 Inc., he works on both the business and the code, from interface design and implementation to deployment and early operations.

He started programming early and learned Python in junior high school. He took on client work through CrowdWorks and Coconala, including web scraping, automation, and small web systems, while also building Twitter bots and operating Minecraft servers and plugins.

While in high school, he worked in restaurant operations, studied in Australia, worked across social media marketing and engineering at stak, served as a full-stack engineer, and ran an independent business. In Australia, he also discussed ideas with local venture investors and worked to launch and develop a new service.

His products include Lina, an AI chatbot for LINE that reached more than 10,000 friends on LINE before the business was transferred; PEOPLE!, a voice-based social app tested for Gen Z; and TRAC, an iPhone app for preserving travel memories. His team also provided AI voice-cloning and avatar-generation technology to a TV Asahi program.

He gives AI agents real work such as research, implementation, review, and operations. His public work includes a context-aware Japanese reading model, AI workflow tools, Cloudflare-based products, mobile applications, open-source utilities, technical benchmarks, and a YouTube channel focused on hands-on technology testing.

Detailed career timeline

Roles, personal responsibilities, outcomes, and sources are listed separately.

Early elementary school–middle school: PC, Python, personal contracting, first web service

I started using PCs in early elementary school, learned Python in middle school, and began taking on web scraping, business automation, and small web-system projects through CrowdWorks and Coconala.

Responsibilities
Web scraping and business automation using Python / Development of Twitter bots / Development and operation of Minecraft servers and plugins / Planning and launching of web services
Outcomes
My first experience using code to deliver value to others / Launched my first web service as a middle school student / Experience in developing and operating services for real users

Self-taught Indie Developer / Sole Proprietor

Source

April 2021 – March 2024: N High School and Student-Facing Services

While attending N High School, I developed an information site for prospective students, a GitBook-based student guide, a GPT-3 learning-support agent, and a student service called Nuixo.

Responsibilities
Planning and implementation of student-facing web services / Operation of an information site and GitBook / Development of a learning support agent using GPT-3 / Design of a low-cost architecture centered around Cloudflare
Outcomes
Initial development and operation of student-facing services / Operated a web service with approximately 4 million monthly pageviews using Cloudflare at a cost of around $5 per month / Pursued automation using generative AI at an early stage

Student / Indie Developer

Source

During high school: Employment at McDonald's

Gained experience in store operations different from technical roles, including customer service, routine tasks, and team collaboration

Responsibilities
Customer service / Store operations / Execution of tasks in a team setting
Outcomes
Experience in on-the-job operations and customer interactions in a non-technical role

Store crew

Source

During high school: Study Abroad in Australia

I experienced daily life and communication in an English-speaking environment in Australia, discussed business ideas with local venture capitalists, and challenged myself by launching and developing a new service.

Responsibilities
Living and studying in an English-speaking environment / Self-management in multicultural environments / Discussing business ideas with local venture capitalists / Design and development of new services
Outcomes
Acquired foundational skills in everyday English conversation, document reading comprehension, and email correspondence / Experience in conceiving new business ventures in overseas environments and turning them into services

International Student

Source

2022: LINE AI Chatbot Lina

I developed Lina, an AI chatbot for LINE, grew it to more than 10,000 friends on LINE, and later transferred the business.

Responsibilities
Planning / Design and implementation of AI chat functionality / Operation and growth on LINE
Outcomes
More than 10,000 friends on LINE / Business transfer

Creator / Developer / Operator / Public source

Source

October 2022–June 2023: Social Media Marketing and Development at stak Inc.

At stak Inc., I worked on social media marketing, new-business PoCs, web development, and event operations. Customer names and individual project details are not public.

Responsibilities
Planning and management of social media content / Development of PoC for new services / Measurement and improvement of marketing strategies / Planning, promotion, and on-site management of events
Outcomes
Multiple records of over 1 million video views / Organized multiple events with over 100 participants / Implemented strategies across marketing and development

Social Media Marketer / Engineer

Source

July 2023–present: Full-stack Development at TSUQREA Inc.

I have been responsible for requirements gathering through front-end, back-end, cloud, and release for multiple web, mobile, and AI projects. Customer names and specific project details are not disclosed.

Responsibilities
Technology selection and requirements definition / Front-end and back-end development for web and mobile / Implementation of AI features and cloud infrastructure / E2E testing, CI/CD, release, and operations
Outcomes
Experience as project lead and tech lead in multiple web, mobile, and AI development projects / Responsible for requirements gathering, implementation, release, and operations throughout the process / Developed PoCs and operational systems incorporating AI and machine learning

Full-stack Engineer / Tech Lead

Source

Student years–present: Sole proprietorship

Since my student years, I have run a sole proprietorship and gained experience in development, business management, contracts, and accounting.

Responsibilities
Operations of a sole proprietorship / Business management / Operational decisions including contracts and accounting
Outcomes
Have run a sole proprietorship since my student years, gaining experience in both development and business operations

Sole proprietor

Source

January 23, 2024–present: Co-founded FP16 Inc.

I co-founded FP16 Inc. during my third year of high school and have been responsible for management, product planning, software development, AI, infrastructure, and operations.

Responsibilities
Management and business decisions / Product planning and UI design / Front-end, back-end, and AI development / Infrastructure, operations, and growth
Outcomes
Co-founded the business on January 23, 2024 / Provided AI technology to a TV Asahi program / Planning and development of new products including PEOPLE!

Co-founder / Representative Director and President / COO

Source

February 2026–present: YouTube Channel 'Ninomiya'

Operating a YouTube channel that conducts real-device verification of Tesla, gadgets, and AI from an engineering perspective, with over 2,800 subscribers and 30 videos published as of July 10, 2026.

Responsibilities
Planning verification themes / Conducting real-device verification and technical evaluation / Production, editing, and publishing / Improvement based on viewer feedback
Outcomes
Over 2,800 subscribers / Published 30 videos / Continuously promoting the practicality of tech, gadgets, and AI

Tech and gadget video creator / Public source

Source

2021–present: Publishing Continuously on GitHub and Zenn

Continuously publishing implementations and insights in web development, AI, local LLMs, and browser tools

Responsibilities
Technical article writing / OSS publication / Sharing evaluation results and implementation processes
Outcomes
Shared implementation insights in Japanese AI, web, and AI agent domains

Technical author / OSS developer / Public source

Source

2026: Japanese AI evaluation and OSS for agents

Published a 2,000-question Japanese RAG benchmark, the Japanese Reading Model, a secure delivery design for AI-generated HTML, and open-source tools for AI agents

Responsibilities
Evaluation data design / Model and API development / Benchmarking / OSS publication and maintenance
Outcomes
100% on 66 difficult-reading questions / 91.9% on 260 independent evaluation questions / Compared six Japanese RAG configurations across 2,000 questions

Research and development / OSS development / Public source

Source

Media and public work

Television, press releases, interviews, podcasts, videos, and talks are recorded with sources.

2024-05-23: TV Asahi — Hayashi Osamu no Ima Shiritai Desho!

Through FP16 Inc., I provided AI voice-cloning and AI avatar-generation technology used in the program.

Provided AI voice-cloning and AI avatar-generation technology / television

Source

2024-07-03: PR TIMES / FP16 Inc. — Opened the waitlist and began testing PEOPLE!, a voice-based social app for Gen Z

I planned and developed PEOPLE!, a voice-based social app for Gen Z built around short voice posts, and launched a public test.

Planning, development, and test operations / press_release

Source

Selected achievements and verification

Grew Lina to more than 10,000 friends on LINE and transferred the business

I developed Lina, an AI chatbot for LINE, grew its user base to over 10,000, and later transferred the business.

More than 10,000 friends on LINE / Business transfer / Developed in October 2022

Source

Independent evaluation of Japanese reading estimation

I developed a Japanese Reading Model combining candidate reading lattices and context rerankers, and compared it against public benchmarks.

100% on 66 difficult-reading questions / 91.9% on 260 independent evaluation questions / The comparison model scored 61.9%

Source

Evaluated Japanese RAG with 2,000 questions

I compared Japanese RAG embedding models using a benchmark of 6 configurations and 2,000 questions.

6 configurations / 2,000 questions / Embedding evaluation for Japanese RAG

Source

Operating at roughly 5 million page views per month

I develop and operate a closed web service portfolio with approximately 5 million monthly page views.

Approximately 5 million monthly page views / Closed web service portfolio

Source

Strengths

Own the full 0-to-1 process end to end

I carry a product from planning and UI through implementation, AI integration, infrastructure, launch, and early operations without splitting the work across silos.

Build small and verify through real-world use

I prioritize a usable prototype within 24–72 hours over extensive up-front design, then use real usage to make decisions.

Use AI as an active agent

I design AI agents as an operational layer for development, research, and operations—not merely as chat assistants.

Design including operations

I treat correctness, maintainability, cost, observability, and user value as equally important design concerns.

Capabilities

AI Agent Design and Operations

Agent system combining tool usage, MCP, memory, RAG, workflows, monitoring, and human approval.

AI agents / MCP / RAG / LLM / tool use / automation

Web Product Development

Full-stack web service development from 0 to 1, including frontend, API, authentication, data, payments, GraphQL, and public infrastructure.

TypeScript / React / Next.js / Hono / NestJS / GraphQL / Stripe / API

Consumer Mobile Product Development

Planning, UI, implementation, App Store release, CI/CD, and feature flags for iOS and Android apps using Expo/React Native.

Expo / React Native / iOS / Android / App Store / feature flags

Practical AI and Machine Learning Systems

Development integrating RAG, real-time speech AI, image anomaly detection, fine-tuning, and GPU inference into web and business systems.

TensorFlow / PyTorch / fine-tuning / OpenAI Realtime API / GPU / RAG

Cloud and Operations Infrastructure

Lightweight architecture centered on Cloudflare Workers, CI/CD, logging, monitoring, and staged deployments.

Cloudflare Workers / D1 / R2 / Durable Objects / AWS / GitHub Actions

Japanese, Speech, and Document Processing

Processing for integrating Japanese into practical systems, including reading estimation, post-speech recognition processing, OCR, and document generation.

Japanese NLP / STT / TTS / OCR / document generation

Product Delivery

Proceeding with requirements gathering, UI/UX, implementation, deployment, automation, and initial user validation as a single workflow.

product design / 0 to 1 / prototyping / operations / growth

Collaboration fit

Public information for deciding whether a project is a good fit for working with Kan.

Good fit

0-to-1 Product Development

Consulting that turns an ambiguous problem into a hands-on prototype, launch, and initial operation in a short time.

Operating AI Agents in Production

Designing AI as an operational layer that handles investigation, implementation, review, and operations, not just as a chat assistant.

Japanese, Speech, AI

Products that utilize Japanese reading, post-STT processing, pre-TTS processing, RAG, and evaluation data.

Collaboration with Shared KPIs

Collaboration that continues after launch and improves usage, retention, revenue, quality, and other shared metrics.

Not a fit

Implementation of Fixed Specifications Only

Requests that only consume effort to fulfill predefined specifications without touching on background or KPIs are not a good fit.

Long-Term Onsite Without Discretion

I do not prioritize roles that only sell my time without room to propose improvements or make technical decisions.

Low-cost, high-frequency support

I prefer products that can be improved through systems rather than services based on manual, individual support.

Large Up-front Builds Without Validation

For long-term development plans without validating the usage hypothesis, I suggest starting with a small validation phase.

Working modes

  • I create a prototype or technical validation that serves as a basis for decision-making within the first 24 to 72 hours.
  • I work primarily remotely, with in-person meetings in Hiroshima adjusted based on the content.
  • I prefer asynchronous collaboration using Issues, Git, short documents, and working prototypes.
  • I welcome messages about collaboration, interviews, speaking, podcasts, video appearances, and open-source work.

Languages and contact

Japanese (native; technical, business, media, and collaboration contexts) / English (Daily Conversation, Document Comprehension, Email, Remote Collaboration)

The easiest way to start is a short email covering the request, desired outcome, deadline, and whether the work can be made public.

Personal preferences and working context

I base my life and development work in Hiroshima and reach users in Japan and abroad directly over the internet, without assuming that I need to relocate.

Preferences

  • I've enjoyed interacting with PCs, software, and gadgets since elementary school.
  • I prefer Mac's consistent development experience and the camera, audio, and physical controls of Sony/Xperia.
  • I'm interested in cars, cameras, audio, smart devices, and local AI.
  • I'm drawn to experiences that let me verify things with my own eyes, such as travel records, the Northern Lights, space, and Shenzhen's technology.
  • I'm also interested in using technology to refine living spaces, such as home theaters and audio systems.

Product taste

  • Products whose value is obvious when used and whose core function is immediately accessible.
  • Fast, accessible, and low-maintenance interfaces.
  • I value clarity as a tool and consistency in the details more than flashy presentation.

AI in daily life

  • Delegate tasks such as research, implementation, testing, review, documentation, and operational verification to AI agents.
  • I use human approval, logs, tests, and diff review to connect AI outputs to real work.
  • Convert thoughts into actionable tasks using voice memos and public context.

Influences

  • Open Web and OSS culture.
  • The belief that individuals and small teams can reach a global audience directly through the internet.
  • Product development that quickly gathers user feedback and iteratively improves.

What I want to build next

This section describes where I am heading, not only what I have done.

Environment where AI agents take on actual work

I turn investigation, development, review, operation, and decision support into continuous operation systems that include observability and human validation.

Foundation for Japanese and Speech Processing

I package reading estimation, speech-recognition post-processing, TTS preprocessing, and evaluation data as reusable infrastructure for other products.

Methods for individuals to operate large products using AI

I am exploring how small teams can sustain development, support, operations, and improvement with AI agents and automation.

Consumer mobile products used every day

I nurture web and mobile services that users want to use daily by spanning planning, development, marketing, and user acquisition.

Build small and deliver directly to the world

I launch internet products with clear purposes quickly and deliver them directly to users in Japan and abroad.

Main technologies

Languages

TypeScript, JavaScript, Python, SQL, Kotlin, Go, Ruby

Frontend

React, Next.js, Vite, Expo, React Native, SvelteKit, Remix, Vue, Tailwind CSS, HTML, CSS

Backend

Node.js, Hono, NestJS, FastAPI, GraphQL, OpenAPI, Prisma, Ruby on Rails, PostgreSQL, MySQL, MariaDB, SQLite, DynamoDB, REST, WebSocket

Cloud

Cloudflare Workers, Cloudflare Pages, D1, R2, Durable Objects, AWS, GCP, Supabase, Vercel, Fly.io

AI and language processing

LLM, AI agents, MCP, RAG, Embedding, OpenAI API, OpenAI Realtime API, TensorFlow, PyTorch, Ollama, LangChain, Whisper, Vertex AI, STT, TTS, OCR

Operations

GitHub Actions, Docker, Playwright, Kubernetes, ArgoCD, Terraform, AWS CDK, CDKTF, observability, automated testing, agent-driven development

Selected writing

Articles that show how I think about implementation, evaluation, and operations.

Interests

  • AI and the Internet
  • Systems for operating AI agents as active contributors
  • Japanese reading, post-speech recognition processing, and pre-TTS processing
  • RAG, Embedding, evaluation data, and search quality
  • A lightweight, easy-to-operate web infrastructure centered around Cloudflare Workers
  • Experiments combining browsers, smart devices, and local AI
  • Internet products that individuals can create quickly and deliver directly to users
  • Growing consumer mobile products from concept through user acquisition
  • Products that preserve experiences by combining travel, movement, maps, photos, and video
  • Using software to reduce real-world waits and constraints in logistics, energy, and mobility

Open to

Please get in touch by email.

Joint development

Joint development of products using AI, Web, voice, and Japanese processing

Product Consultation

Consultation on new product development, technology selection, AI integration, and verification planning.

Interviews & Media

Interviews about AI agents, indie development, Japanese AI, young entrepreneurs, and product development.

Speaking & Technical Outreach

Speaking engagements and study sessions where insights from implementation, evaluation, and operation are shared.

OSS Collaboration

Contributing Issues, PRs, and collaborative verification to publicly available tools and agent foundations.

kan@2-38.com

Personal notes

  • 20 years old (as of July 10, 2026).
  • I grew up in Hiroshima and am still based there.
  • My personality type is ENFP.
  • I enjoy AI and the internet.
  • I have sloping shoulders.

External links

  • GitHubhttps://github.com/kandotrun
  • X / Twitterhttps://x.com/kandotrun
  • Zennhttps://zenn.dev/nixo
  • YouTubehttps://www.youtube.com/@nixo0
  • Wantedlyhttps://www.wantedly.com/id/kan_ninomiya
  • YOUTRUSThttps://youtrust.jp/users/c7cf664c8adb63f086d67016a2427889
  • 2-38.comhttps://2-38.com/
  • Emailmailto:kan@2-38.com