友田 陽大
日本語English
Developer of a METI Minister's Award–winning product

SaaS / industry DX —from architecture to infrastructure,delivered single-handedly

B2B SaaS, legacy-industry DX, and production-grade generative AI (RAG) on TypeScript + Python + AWS, delivered solo from requirements through infrastructure and operations. Powered by the latest generative AI (Claude Code), end to end across implementation, testing, E2E, and IaC. Delivered fast, cost-effectively, and securely.

Book a free consultSee the award-winning work
  • METI Minister's Award–winning product
  • 3+ services in production, delivered solo
  • NDA before any call · copyright assigned to you

Proof & standards

  • METI Minister's Award

    Award-winning product

  • Kyoto-certified

    Industry DX system

  • OWASP Top 10

    B2B SaaS in production

  • Terraform IaC

    Cloud-native architecture

Track record (numbers)

  • 0

    METI Minister's Award–winning product

  • 0

    Products built & shipped

  • 0%

    End-to-end (one-stop) delivery

  • 0+

    Core technologies in the stack

External Recognition

CrowdWorks contract rankings — top results

Contract-ranking results on CrowdWorks, one of Japan's largest crowdsourcing platforms. Each result is verifiable on its public ranking page.

  • #1

    Engineer division

    Weekly contract ranking

    Feb 2026 · Week 1

  • #1

    Overall

    Weekly contract ranking

    Feb 2026 · Week 1

    View ranking

  • #4

    Engineer division

    Monthly contract ranking

    Feb 2026

    View ranking

  • #6

    Overall

    Monthly contract ranking

    Feb 2026

    View ranking

Problem

Do any of these sound familiar?

The “delivery wall” many companies hit. Every one of these is a problem I've solved single-handedly.

  • You have the idea, but no engineering team

    Common at startups and new ventures: no in-house engineering, and splitting the work across separate vendors for design, frontend, backend, and infrastructure makes coordination costs balloon.

  • You want DX, but it's too analog to even start

    Workflows run on Excel, fax, and phone calls. You know you need DX, but you can't find a developer who can systemize your industry's complex operations — and big SIers are expensive and inflexible.

  • You adopted AI, but it's stuck at PoC

    You can call the ChatGPT API, but it never reaches production quality. RAG accuracy, response latency, and AWS architecture can't keep up — you can't cross the wall from “research” to “real use.”

I handle every one of these single-handedly.

AI-Augmented Delivery

Generative AI × solo full-stack — fast, cost-effective, secure

From requirements to implementation, testing, E2E, IaC, and operations — delivered solo with the latest generative AI (Claude Code). AI is the accelerator; quality is guaranteed by human verification gates.

  1. Requirements

    Sharpen the problem

  2. Design

    Types, boundaries, data

  3. Build (AI pair)

    Accelerated with Claude Code

  4. Tests & E2E

    Unit → API/E2E automated

  5. IaC

    Reproducible with Terraform

  6. Production

    Monitoring, resilience, idempotency

  • Fast

    Fast

    Pair-building with generative AI (Claude Code) speeds up implementation, tests, and refactoring. I'm the single point of contact from design to infrastructure, so there's zero multi-vendor coordination lag — the shortest path to an MVP.

  • Cost

    Cost-effective

    The hours AI saves go straight back into the rate. Serverless on Lambda, DynamoDB, and Vercel plus Terraform IaC means production from a few hundred yen a month at first, with low maintenance cost.

  • Safe

    Secure

    Even AI-written code always passes human verification gates — type checking, automated tests, E2E, and dependency/vulnerability scanning (OWASP Top 10). Copyright is assigned to you, and an NDA can be signed before any call. Speed and quality are not a trade-off.

“Is AI-built quality really OK?” — Here's the proof, in numbers.

  • 2,153

    Automated tests (backend)

  • 96.7%

    Type coverage (another project)

  • 0

    Missing-authorization findings (4 audits)

  • 0

    Double charges in production (payments platform)

Solution

Five strengths that solve the problem

From requirements to infrastructure — speed, quality, and security, with no compromises.

  • Fastest MVP

    One-stop, zero-to-one delivery

    Everything by one person, from design to infrastructure. No multi-vendor coordination — the shortest path to an MVP release. Ideal for a startup's race against time.

    Example

    Requirements · design · React · Go/Python · AWS/Terraform

  • METI Minister's Award

    Analog-industry DX

    Turning Excel/fax/phone workflows into full SaaS. Built a METI Minister's Award– and Kyoto-certified system for the lumber industry. I capture an industry's tacit knowledge at high resolution and translate it into design.

    Example

    Lumber-industry DX: Excel/fax ops → full SaaS →
  • 90%+ accuracy

    Production-grade AI / RAG

    AI that doesn't stop at a PoC. RAG (pgvector) with AWS Bedrock delivers production-level accuracy and response speed, including UX gains through parallelization.

    Example

    AI voice chatbot (AWS Bedrock + pgvector) →
  • 70% faster response

    High-performance design

    Concurrency in Python/Node.js speeds up I/O-bound work. Tuning voice and batch processing sharply improves perceived response time.

    Example

    AI voice chatbot: parallelized audio processing, 5s → 1.5s →
  • 90%+ satisfaction

    Intuitive UX design

    Visual UI that crosses language barriers. Swipe UIs and i18n make interfaces anyone can use without getting lost.

    Example

    Restaurant matching: 90% satisfaction via an intuitive UI →

Case Studies

Selected work

Analog-industry DX, production AI, high-speed matching systems — a range of projects, each delivered solo.

View all case studies

Indie Products

Validating contract-honed engineering in my own SaaS

The architecture, SEO, and payments patterns honed in client work get re-sharpened against real operational load by running my own indie SaaS under a single brand.

  • Hakokit

    Everyday work tools, built by an indie developer

    Launched Apr 2026

    An indie SaaS portfolio for re-validating the architecture, SEO, and payments patterns honed in client work against my own products. Several services run under one brand, with shared foundations (auth, billing, AI Gateway) in a hexagonal monorepo — an experiment in an operating model where a new app can be added in a matter of days.

    • A Next.js 16 + Turborepo monorepo running several SaaS products from one place
    • Centralized token-cost control via an AI Gateway + prompt cache
    • Automatic Japanese consumption-tax calculation and qualified-invoice issuance with Stripe Tax
    • A marketing foundation with programmatic SEO / GEO (llms.txt) as standard
    Next.js 16
    React 19
    TypeScript
    Turborepo
    Supabase
    Stripe
    Anthropic API
    Tailwind CSS
    Cloudflare Workers

  • Aegis

    A defense-in-depth security toolkit for Next.js / Supabase SaaS

    Launched Jun 2026

    The security-implementation patterns honed across client and product work, crystallized into a drop-in OSS + CLI. A single middleware file and typed env automate the horizontal controls (headers/CSP, rate limiting, CSRF, secrets hygiene), while it detects and flags the vertical risks a library can't fix (authorization/IDOR, Supabase RLS misconfigurations) via taint analysis and SQL validation. MIT-licensed.

    • One middleware file + typed env automate the horizontal controls vibe coding misses
    • Intraprocedural taint analysis detects SQLi / SSRF / IDOR with source→sink traces
    • Validates RLS / SQL in supabase/migrations and correlates with code to confirm exposure
    • Safe auto-fixes + SARIF for GitHub code scanning (only high-confidence findings block CI)
    TypeScript
    Next.js
    Supabase
    PostgreSQL RLS
    Zod
    SARIF
    Vitest

Stack

Modern technology, robust systems

Type safety with TypeScript and Zod, IaC with Terraform, and implementation-through-E2E automation powered by generative AI (Claude Code) — systems built to last, fast and secure.

  • Frontend

    • React 19
    • Next.js (App Router)
    • TypeScript
    • Vite
    • TanStack Query
    • Zod
    • Tailwind

  • Backend

    • Go (Echo, wire)
    • Python (Flask, FastAPI)
    • Node.js
    • SQLAlchemy

  • Infrastructure

    • AWS
    • GCP
    • Terraform
    • ECS / Fargate
    • Lambda
    • Vercel
    • Docker

  • Data

    • PostgreSQL
    • pgvector
    • DynamoDB
    • Redis
    • Supabase

  • AI / LLM

    • AWS Bedrock
    • Claude API
    • RAG
    • LangChain
    • Whisper
    • Polly

  • AI-driven dev / QA

    • Claude Code
    • Vitest / pytest
    • Playwright (E2E)
    • GitHub Actions
    • ESLint / Prettier
    • OWASP ZAP
    • npm / pip audit
    • Sentry

Turn your conceptinto a working product

“Is it feasible?” “Need an MVP fast?” “Want advice from tech selection on?” — start with a free 30-minute consult.

Book a free consultSee pricing
  • NDA available
  • Fully remote
  • Quasi-mandate, contract & advisory
  • Invoicing & contracts handled