Author Image

こんにちは、 Clarence

Clarence

Senior Backend Developer / LLM & Infrastructure Engineer at BRONCI TECHNOLOGY INC.

Backend、AI Application、DevOps、Infrastructure を中心に、API、CI/CD、Cloud、On-prem、Network、Security をひとつの delivery chain として扱っています。単に機能を作るだけではなく、サービスを安定して公開し、観測でき、運用できる状態にすることを重視しています。必要に応じて LLM、Speech AI、GPU Runtime、Cloud Services、自社運用の infrastructure を実際のプロダクトに統合します。現在は長問科技で Senior Backend Developer / LLM Engineer として、Backend service、API、GitLab CI/CD、GitHub Actions、AWS CDK / IaC、Kaldi、Whisper、OpenAI / Ollama、Proxmox VE、Harbor、LibreNMS、LDAP、Enterprise Wi-Fi / SD-WAN、機材選定を担当しています。AWS CDK 完全學習手冊 の著者でもあり、Cloud、DevOps、Linux、Network、AI 開発の経験を記事や Open Source として整理しています。

スキル

AWS CDK
AWS CDK

Infrastructure as Code (IaC) の実務経験があり、AWS CDK 完全學習手冊 の著者です。TypeScript / Python で Cloud infrastructure を管理します。

経歴

1

Taipei

Backend、AI Application、Cloud / On-prem Infrastructure、DevOps automation を担当

Senior Backend Developer / LLM Engineer

11 2020 - 現在

担当:
  • Node.js、Python、Go、C/C++、Shell Script、Java で Backend service と internal tools を開発。
  • API、background jobs、data processing flow、third-party service integration を設計。
  • 日本向け home camera cloud service を統合し、remote monitoring と device management を支援。
  • 日本向け IoT smart home voice command service を開発。
  • 台湾向け home appliance voice-control solution を device、cloud、voice workflow まで接続。
  • Jest automated tests と regression checks を整備し、backend release の再現性を維持。
  • 2022 年から LLM application、tool calling、agent workflow、production integration に取り組む。
  • OpenAI API、Ollama、internal AI workflow を product / operations use case に統合。
  • Kaldi、Whisper speech recognition を統合し、ASR、transcription、model deployment を担当。
  • NVIDIA GPU server、driver / runtime、inference environment、system tuning を運用。
  • AWS architecture を availability、scalability、networking、deployment path の観点で設計。
  • AWS CDK / Infrastructure as Code で再現可能な cloud resources を管理。
  • AWS、GCP、Azure、On-prem services を接続し、cloud と self-hosted systems を連携。
  • GitLab CI/CD、GitHub Actions、container image build、deployment workflow を維持。
  • On-prem、cloud、physical data center の運用、hardware maintenance、capacity planning を担当。
  • Proxmox VE cluster、virtualization、storage、VM lifecycle を設計・運用。
  • Proxmox Backup Server を構築し、backup、restore、disaster recovery workflow を整備。
  • Harbor private registry、GitLab、LDAP など internal platforms を導入・維持。
  • Office / enterprise wired network、VLAN、routing、firewall policy、switching を計画。
  • SD-WAN distributed network を管理し、cross-site connectivity、performance、security を支援。
  • Aruba、UniFi Wi-Fi を構築し、office wireless coverage、roaming、operations を担当。
  • LibreNMS で network / server monitoring、alerting、issue detection を実施。
  • Internal / external data-center path、site-to-site backup、high availability を計画。
  • Dell、HP、Supermicro server hardware と GPU server specification を評価。
  • Switch、firewall、wireless AP など network equipment を評価。
  • Vendor quote と solution comparison を整理し、保守可能な purchase specification に落とし込む。

Taipei / Taichung / Kaohsiung

Network Planning / Deployment

2024 - 2024

担当:
  • Office and store network planning、wiring、Wi-Fi、network device deployment を支援。
2

3
Business Next

06 2022 - 現在

Writer

Column Writer

06 2022 - 現在


DrMaster Press

11 2021 - 現在

Author

AWS CDK 完全學習手冊:打造雲端基礎架構程式碼 IaC

Book Author

11 2021 - 現在

担当:
  • AWS CDK を基礎から解説する中国語書籍を執筆。
  • 豊富な sample code を用意し、IaC の実践を学びやすく整理。
4

5

Taipei

Cloud system software research and development

Full-time Research Assistant

11 2015 - 11 2020

担当:
  • Network application development - Node.js, Go, C/C++, PostgreSQL
  • Backend development - Node.js, PostgreSQL, Docker

Volunteer

SITCON Credits の公開記録に基づく年次カンファレンス運営スタッフ履歴

SITCON 2018 製播組メンバー

2018 - 2018

SITCON 2017 行銷組メンバー

2017 - 2017

SITCON 2016 行政組チケット担当・行銷組メンバー

2016 - 2016

SITCON 2015 行銷副組長

2015 - 2015

SITCON 2014 場務組メンバー

2014 - 2014

6

7
SITCON Summer Camp

2014 - 2015

Volunteer

SITCON Credits の公開記録に基づくキャンプ運営スタッフ履歴

SITCON Summer Camp 2015 - 副召

2015 - 2015

SITCON Summer Camp 2014 - 課程組長

2014 - 2014

学歴

M.S. Electrical Engineering
Higher Vocational School, Information Technology

プロジェクト

clarence.tw
clarence.tw
Owner Current

Hugo Toha で構築した multilingual personal site。Backend、AI/LLM、DevOps、Cloud / On-prem Infrastructure、Network Operations を整理しています。

Clarence Tech Blog
Clarence Tech Blog
Author Current

LINE、AWS、Ubuntu / Linux、Proxmox VE、Docker、Node.js、GitLab、Network、NVIDIA、troubleshooting notes をまとめる技術ナレッジベース。

LLM Text Summarization Platform
LLM Text Summarization Platform
Builder 2022 - Present

OpenAI API、Ollama、local models、batch processing を統合した internal Text Summarization / AI application platform。

Speech AI and GPU Runtime
Speech AI and GPU Runtime
Builder / Operator 2022 - Present

Kaldi、Whisper、GPU runtime を統合し、Speech AI、model inference、system deployment を支援。

Self-hosted GitLab, Runner and Harbor
Self-hosted GitLab, Runner and Harbor
Operator 2022 - Present

Self-hosted GitLab、GitLab Runner、Harbor registry、CI/CD workflow を運用し、source code、container image、deployment を接続。

Proxmox VE / PBS Infrastructure
Proxmox VE / PBS Infrastructure
Operator 2023 - Present

Proxmox VE、Proxmox Backup Server、GPU server、On-prem Infrastructure を運用し、internal services、AI runtime workload、backup strategy を支援。

High-traffic Web Hosting Operations
High-traffic Web Hosting Operations
Operator 2023 - Present

High-traffic hosting environment と多数の Web サイトを長期運用し、monitoring、updates、capacity planning、security maintenance を担当。

node-crypto-ecb
Owner Sep 2022 - Present

Compatibility-driven な encryption / decryption のための Node.js AES ECB library。

cdk-wordpress
Owner Oct 2020 - Present

WordPress deployment を AWS CDK Construct Library として再現可能な IaC にしたプロジェクト。

出版

AWS CDK 完全學習手冊:打造雲端基礎架構程式碼 IaC(iT邦幫忙鐵人賽系列書)

AWS CDK を基礎から学ぶための中国語書籍。豊富な sample code を通して Infrastructure as Code を実践できます。

注目記事

私の技術的な仕事の進め方:Backend、AI、DevOps、Infrastructure

私は Backend、AI、DevOps、Infrastructure、Network を別々の領域としてではなく、実際に動くプロダクトを支えるひとつの delivery chain として扱っています。 機能を使える状態にするには API が安定している必要があります。サービスを更新するには CI/CD が必要です。長期運用するには monitoring、backup、network、security が必要です。LLM、Speech AI、GPU runtime を使う場合も、model integration は deployable、observable、operable でなければなりません。 主に扱う領域 Backend / API Backend は私の中心的な領域です。API design、service integration、database access、既存システム連携、admin workflow を扱います。Node.js、Python、Go、PHP、C/C++、Shell Script、Java を、既存アーキテクチャや運用性に合わせて使います。 重視しているのは、API が明確であること、data flow を追跡できること、error を診断できること、そして次の担当者が保守できることです。 LLM / Speech AI 2022 年から LLM と Speech AI を実際のプロダクト workflow に組み込んでいます。単に model API を呼ぶだけではなく、prompt、batch processing、local model、GPU runtime、OpenAI API、Ollama、Kaldi、Whisper、既存 backend workflow を接続します。 入力データの整理、出力の検証、retry、cost control、local model と cloud API の選択、runtime environment の再現性を重視します。