I’m drawn to products that improve how people work, and to products that can become infrastructure for society. If people can spend their time on meaningful, interesting work and have their individual capabilities empowered, I think more of us will actually enjoy working.
At the same time, I’m fascinated by how technology has historically rewritten how society itself works — payment infrastructure, Web3, generative AI — and I want to keep working on the side that treats technology not as a mere tool but as a force that reshapes society.
Education
The University of Tokyo — M.Eng. in Precision Engineering (Apr. 2026 – Mar. 2028)
The University of Tokyo — B.Eng. in Precision Engineering (Apr. 2022 – Mar. 2026)
Work Experience
Sansan Inc.— Full-Stack Engineer (Internship)
Feb. 2026 – Mar. 2026
Redesigned the async integration that handles bulk business-card distribution. Replaced an in-progress design with a Pub/Sub × Cloud Tasks BufferTask hybrid that separates fan-out from execution-side rate control, removing dead-letter failures caused by upstream API rate limits and the manual recovery work that used to follow them.
Cloud Tasks Pub/Sub GCP Node.js Async
Bizgem Inc.— Full-Stack Engineer (Internship)
Sep. 2025 – Present
Built API and RPA integrations connecting major e-commerce platforms (Amazon, Rakuten, Shopify, ZOZO, Next Engine) to a SaaS that consolidates inventory and sales data for retailers. Worked across the stack — from an async batch pipeline on AWS Lambda + Celery + Redis, to AI-assisted development infrastructure, to improvements to the GitHub Actions CI.
AWS Aurora Lambda Celery Redis Next.js GitHub Actions
AICE Inc.— Software Engineer Intern
Dec. 2023 – Apr. 2025
Drove PoC projects end-to-end as the engineering lead of a 3-person team, while also contributing as an engineer to enterprise projects. In a joint project with a railway company, built a generative-AI–driven recommendation product that delivered a new experience for restaurant and tourism discovery; on the enterprise side, built RAG-based document search for a corporate chatbot.
GCP Cloud Run Vertex AI GPT-4o RAG Terraform TypeScript
Tech Skills
- Languages:
- Frameworks:
- Cloud:
- Tools:
- Spoken:Japanese (Native) · English (TOEFL iBT 84)
Publications
AWS Auroraを論文から読み解いてみる ↗
[Zenn]2026-04-23
Redesigning the async fan-out infrastructure for Sansan Bizcard Maker ↗
[Sansan Builders Box]2026-03-31
How we replaced a Cloud Tasks-only design with a Pub/Sub × Cloud Tasks BufferTask hybrid to handle 3,000+ card bulk distributions without dead-letter failures.
GCP Pub/Sub Cloud Tasks Async
【Python】文字列連結のパフォーマンス最適化 — joinって本当に速いの？ ↗
[Zenn]2025-12-22
Personal Projects
django-coupling ↗
A static coupling-analysis CLI for Django / Python projects — a Python take on cargo-coupling. Scores Strength, Distance, and Volatility from AST usage and git history, and flags layer violations where a lower layer (models) is imported by a higher one against the expected views → services → models direction.
Python Django Static Analysis CLI
p2pboard ↗
A serverless P2P whiteboard / chat CLI that runs over a local LAN or a Docker network. Peers in the same room discover each other automatically via UDP multicast, and messages are delivered over persistent TCP connections. Humans use it from a terminal; AI agents join the same room through a JSONL input/output interface. Low-level socket / bind / setsockopt handling written directly with `golang.org/x/sys/unix`.
Go P2P CLI UDP
minify ↗
A compact workflow builder that composes PDF processing, text formatting, and LLM inference as DAG nodes. React-based node editor + async FastAPI backend, with unified switching across OpenAI / Anthropic / Gemini.
LLM FastAPI React DAG