About

I am a backend-focused engineer who carries work through requirements, design, implementation, testing, release, and operations. Most of it is Python and TypeScript, building systems that move the business forward.

What I am usually brought in for is choosing the technology and architecture, and designing for availability and reliability. I migrated an existing DRF-based API to Django Ninja + Pydantic, making the schema the single source so that type safety and the API contract stayed consistent. I built exclusive control for Celery jobs firing every ten minutes in a multi-tenant environment, so the same task never runs twice across tenants — implemented as a shared mechanism other jobs could adopt. In both cases the time went into making the problem not happen twice, rather than into getting something to work.

On the front end I also work in React / TypeScript; for infrastructure I build and run AWS (ECS Fargate, Lambda, RDS, S3) with Docker and Terraform. On the AI side I designed a feature that embeds Amazon Bedrock in a production system, generating CSV join plans as structured output against a JSON Schema. I also led the architecture design of a RAG system and worked on a demand-forecasting PoC with time-series models.

I am easiest to bring in for rebuilding or type-hardening an existing API, designing a practical way to put generative AI inside a business system, improving the performance of data processing, and projects that need one person to stand things up including the infrastructure. It is fine if the requirements are not settled yet — I can start by putting together the material for the decision.

Experience

2021 — present, 11 projects. Backend-focused, working as developer, lead, and sub-lead.

Focus areas

Backend development, API design, data platforms, generative AI, local environments with Docker, and analytics in SQL

Core technologies

Python, TypeScript, Docker, Terraform, SQL

Where I add most

Evaluating and introducing new technology, debugging (reproducing, isolating, and finding root causes), database design, and consolidating duplicated implementations into shared code

Certifications

JDLA Deep Learning for ENGINEER / Fundamental Information Technology Engineer (Japan) / AWS Developer Associate / AWS Solutions Architect Associate

Education

National university (regional), Faculty of Engineering — Electrical Engineering

Traits

Lazy in the engineering sense (never doing the same work twice) / strongly curious

In my own words

There are more situations than ever where failing fast and correcting course works well. What raises the accuracy of a decision, though, is understanding the domain — the deeper that understanding, the less rework, and the faster things go as a result. Rather than treating speed and understanding as opposites, I invest in understanding and take the long-term speed that comes with it.

CONTACT

Send me two or three lines about where you are and what is blocking you, and I will come back within five business days with feasibility, a rough estimate, and how I would approach it. It is fine if the requirements are not settled yet.

contact@yomaru-tech.com