# Kos 的完整履歷

# Kos's full resume

Canonical: https://www.aikos.cc/resume/

專注金流可靠性、帳務一致性與異常復原。曾任支付、錢包與內容平台技術負責人，帶領 13 人以上團隊，以冪等、分散式鎖與自動對帳避免重複扣款、確保帳務一致；現以 Laravel、Golang、Python 與 Node.js 建構 AI Agent 工作流程與後端服務。

## 正在找新機會
- 職位: 資深後端工程師、技術主管
- 型態: 全職、約聘、接案
- 地點: 台灣、海外或遠端
- 語言: 中文和英文

## 序章 (2008–2013): 從客戶的需求開始寫起

2008 年在知識科技入行，同時扛好幾個企業網站、活動網站與客製化平台。之後用 ASP.NET 與 C# 寫企業網站與內部管理系統；在易透網，用 Delphi 寫 ActiveX 接上晶片卡讀卡機，為政府機關與企業做身分驗證與安全登入。

## 第一章 (2014–2018): 第一次帶人，第一次對帳

在幸福家不動產擔任資訊主管，帶 4 人以下的團隊，從架構、開發到 Linux 部署都自己來。接著在聲聯網重構多層通路分佣與帳務結算系統：可配置的分佣規則引擎、帳務流水與資金稽核。從這裡開始，算錢這件事不能出錯。

## 第二章 (2019–2024): 規模與速度

在傲創科技從零建置新聞採集平台與全文搜尋系統，每日處理十萬級新聞資料。可配置式採集框架讓新站點接入從約 2 天縮到半天內；搜尋回應從約 2 秒優化到 300 毫秒內。

## 第三章 (2024–2026): 一筆錢只能入帳一次

擔任支付、錢包與內容平台的技術負責人，管理 13 人以上團隊。回呼重送、網路重試都可能讓同一筆款項被處理兩次，於是用冪等、分散式鎖與自動對帳，讓同一筆交易只入帳一次，會員餘額與資金流水隨時對得上。

## 現在: AI Agents

現在以 Laravel、Golang、Python 與 Node.js 建構 AI Agent 工作流程與後端服務。工作中也把 AI 用在實處：分析分佣規則、檢測帳務異常、解析文件，以及輔助開發與監控。下一章，正在找。

18 年軟體開發經驗，待過 7 間公司、做過 9 個代表專案。

## 2024/11 – 2026/05 · 資訊主管, 公司名暫不公開

- 軟體專案主管，管理 13 人以上

技術: Kafka, Redis, MongoDB, MySQL, API Gateway, Microservices, WebSocket, CDN

### 支付系統
- 擔任支付系統技術負責人，規劃系統架構與技術方案
- 安排團隊開發工作、Code Review 及技術指導
- 協調產品、客服、風控、財務等跨部門需求與開發時程
- 開發支付訂單、充值、提款及對帳系統
- 建立支付冪等、補償、重試及風控機制
- 導入 Kafka、Redis、MongoDB 建立事件驅動架構
- 優化 MySQL 查詢效能及高併發交易流程
- 建立交易監控、告警及自動化維運流程

### 錢包系統
- 擔任錢包系統技術負責人，規劃高併發資金交易架構
- 管理開發人員工作分配與專案進度追蹤
- 協調支付、財務、風控及測試團隊完成系統整合
- 開發會員錢包、帳務及資金流水系統
- 建立餘額一致性、交易冪等及分散式鎖機制
- 使用 Redis Cache 提升高併發交易效能
- 建立自動對帳、補償及風控稽核流程
- 規劃高可用錢包服務架構

### 內容服務平台
- 主導內容平台架構設計與開發規劃
- 安排團隊開發工作並協調第三方廠商串接
- 協調前端、後端、QA 及維運團隊共同完成專案
- 開發內容平台前後台及第三方內容服務串接
- 建立 API Gateway 與微服務架構
- 導入 Redis、CDN、WebSocket 提升系統效能
- 優化 API、快取及高流量存取機制
- 建立監控、告警及 AI 輔助開發流程

## 2019/01 – 2024/10 · PHP 資深工程師, 傲創科技股份有限公司

- 網際網路相關業，100 至 500 人

技術: PHP, Hyperf, Swoole, RabbitMQ, Redis, APCu, MongoDB, Elasticsearch, Canal, MySQL

- 新站點接入時間: 約 2 天 → 半天內
- 搜尋回應時間: 約 2 秒 → 300 毫秒內
- 每日處理新聞資料: 十萬級

### 新聞採集平台與全文搜尋系統
- 從零建置新聞採集平台與全文搜尋系統
- 使用 Hyperf、Swoole 建立高併發新聞採集服務
- 導入 RabbitMQ 建立任務調度、削峰填谷及非同步處理機制
- 建立 Redis、APCu 多層快取架構，提升系統效能與降低資料庫負載
- 使用 MongoDB 管理採集規則及非結構化資料
- 建立 Elasticsearch 全文搜尋平台及索引管理機制
- 整合 Canal 完成 MySQL 即時同步至 Elasticsearch
- 建立採集規則管理、失敗重試、例外處理及監控機制
- 建立可配置式採集框架，提升新站點導入效率
- 建立日誌、監控及告警機制，提升系統穩定性

成果
- 從零建置高併發新聞採集與全文搜尋平台
- 每日穩定處理十萬級新聞資料
- 新站點接入時間由約 2 天縮短至半天內
- 搜尋回應時間由約 2 秒優化至 300 毫秒內
- 建立可擴充採集框架，提升後續開發效率
- 建立高可用、高擴充性的全文搜尋架構

## 2016/10 – 2018/01 · 資深研發工程師, 聲聯網股份有限公司

- 電腦系統整合服務業，1 至 30 人
- 台北市中正區

技術: RabbitMQ, Redis, Account Ledger, AI

### 通路分佣與帳務結算系統
- 重構多層通路分佣與帳務結算系統
- 建立可配置式分佣規則引擎，支援多層代理、多通路及階梯式分佣規則
- 開發佣金計算、批次結算及帳務處理流程
- 導入 RabbitMQ 建立非同步批量結算及任務調度機制
- 建立 Account Ledger（帳務流水）及資金稽核機制
- 建立 Redis 快取機制，降低重複運算及提升結算效能
- 建立交易監控、異常補償、重試及對帳機制
- 優化結算、對帳及帳務流程，提升系統穩定性與擴充能力
- 導入 AI 協助分佣規則分析、異常檢測及帳務流程優化

成果
- 重構多層通路分佣與帳務結算系統
- 建立可配置式分佣規則引擎，支援多通路、多層級及自訂分佣策略
- 導入 RabbitMQ 非同步批量結算機制，提升結算吞吐量
- 建立完整 Account Ledger 與資金稽核流程，提升帳務一致性
- 分佣規則可透過後台設定，降低維護成本並提升需求交付效率
- 建立高可用、高擴充性的帳務與結算架構

## 2014/08 – 2016/10 · 資訊主管, 幸福家不動產

- 軟體專案主管，管理 4 人以下
- 不動產經營業，100 至 500 人
- 台南市中西區

技術: PHP, Python, MySQL, RESTful API, Composer, S3, Linux, Nginx

### 企業管理系統
- 擔任技術主管，規劃系統架構、技術選型及開發流程
- 安排團隊開發工作、工作分配及 Code Review，協調產品、業務、QA 及其他部門共同完成專案
- 主導企業核心平台架構設計與核心模組開發
- 使用 PHP、Python 開發企業內部管理系統及後端服務
- 設計與開發 RESTful API，整合內外部系統資料交換
- 規劃 MySQL 資料庫架構、索引設計及 SQL 效能優化
- 導入 Composer 套件管理及標準化開發流程
- 建置 PDF 文件解析、自動匯入及資料清洗系統
- 開發資料採集、爬蟲及資料同步服務
- 規劃檔案管理架構，整合 S3 Object Storage 檔案儲存服務
- 建立檔案上傳、下載及權限管理 API
- 負責 Linux、Nginx 環境建置、系統部署及版本發布
- 建立系統監控、維運及異常排除流程
- 導入 AI 協助文件解析、資料分析及系統效能優化

成果
- 建立高可用企業管理平台及後端服務架構
- 完成內外部系統 API 整合，提升資料交換效率
- 建立 PDF 自動解析及資料匯入流程，降低人工處理成本
- 建立自動化資料採集與同步機制，提升資料即時性
- 建立集中式檔案管理平台，提升檔案管理效率
- 建立標準化開發流程，提升團隊開發效率與系統維護性

## 2011/10 – 2013/10 · 工程師, 易透網股份有限公司

- 電腦軟體服務業，1 至 30 人
- 台南市中西區

技術: PHP, Delphi, ActiveX, MySQL, JavaScript

### 晶片卡身分驗證系統
- 參與政府機關及企業內部資訊系統開發與維護
- 與客戶進行需求訪談、需求分析及功能規劃
- 協助客戶釐清業務流程，提出系統設計與技術解決方案
- 與專案經理、客戶及開發團隊協調需求、時程及功能驗收
- 負責需求分析、系統設計、功能開發、測試、部署及維運
- 使用 Delphi 開發 ActiveX 控制元件，整合晶片卡讀卡設備
- 建立晶片卡憑證讀取、身分驗證及安全登入機制
- 使用 PHP 開發身分驗證平台及後端服務
- 整合 ActiveX 與 PHP 後端，完成登入授權及會員管理系統
- 建立會員資料、權限管理及登入紀錄機制
- 開發 RESTful API，整合前後端及相關系統資料
- 處理瀏覽器、讀卡設備及後端系統整合問題
- 負責系統測試、問題排除及技術文件撰寫
- 優化登入流程及提升系統穩定性與安全性

成果
- 參與政府機關及企業資訊系統建置
- 建立晶片卡身分驗證及安全登入平台
- 完成 Delphi ActiveX 與晶片卡讀卡設備整合
- 建立 PHP 後端登入授權、會員管理及權限控管系統
- 提升瀏覽器與讀卡設備相容性及登入穩定性
- 協助客戶完成需求導入及系統上線驗收

## 2010/10 – 2011/10 · 工程師, 小兵立大功股份有限公司

- 電腦軟體服務業，1 至 30 人
- 台南市北區

技術: ASP.NET, C#, SQL Server, JavaScript

### 企業網站與內部管理系統
- 負責企業網站與內部管理系統開發維護，參與需求分析、功能設計、程式開發及上線維運
- 使用 ASP.NET 與 C# 開發後端商業邏輯，建置會員管理、資料維護、查詢及報表等核心功能模組
- 整合 SQL Server 資料庫，負責資料存取、查詢優化與資料正確性驗證，確保系統效能與穩定性
- 負責系統功能測試、問題排查與版本更新，確保系統穩定運行並滿足業務需求

## 2008/02 – 2010/10 · 工程師, 知識科技股份有限公司

- 網際網路相關業，30 至 100 人
- 台南市中西區

技術: Third-party APIs

### 企業網站、活動網站及客製化平台
- 負責企業網站、活動網站及客製化平台系統開發，涵蓋需求分析、系統設計、功能開發、測試、部署與後續維護，提供完整專案交付
- 依據需求規劃系統功能與資料流程，完成資料庫設計、後端商業邏輯開發、前後端整合及版型建置，交付符合業務需求的客製化系統
- 開發並整合第三方 API 與外部系統服務，提升系統整合能力與資料交換效率
- 同時負責多個專案開發與維護，依時程控管開發進度，確保系統品質與如期交付
- 協助進行問題分析、系統優化、功能擴充與技術支援，持續提升系統穩定性

成果
- 獨立完成多個企業網站、活動網站及後台管理系統建置，涵蓋會員系統、問卷系統、形象網站、活動平台及企業管理系統等類型
- 建立可重複使用的功能模組與開發流程，提升客製化專案開發效率，縮短需求開發與交付時間
- 成功整合多項第三方服務與 API，提升系統擴充性，降低後續維護成本
- 在多專案並行情況下，維持專案品質與交付時程，累積跨產業系統開發與客製化解決方案經驗

## 支付與錢包系統

技術負責人・帶領 13 人以上團隊・2024–2026

高併發交易下，回呼重送或網路重試都可能讓同一筆款項被處理兩次；會員餘額與資金流水必須隨時對得上。

- 支付冪等、分散式鎖與餘額一致性機制，同一筆交易只入帳一次
- 補償與重試機制處理失敗交易，搭配風控稽核
- 以 Kafka、Redis、MongoDB 建立事件驅動架構
- 自動對帳流程，加上交易監控與告警
- 以重送、併發與補償情境的測試驗證上述機制

- 不重複扣款: 冪等與分散式鎖
- 自動對帳: 補償與稽核流程

公司名稱與營運數據暫不公開。可公開的是測試情境：回呼重送、併發請求與補償，沒有另附線上交易量。

## 新聞採集與全文搜尋平台

傲創科技・資深工程師・2019–2024

每接一個新聞網站就要寫一套爬蟲，約需 2 天；全文搜尋回應約需 2 秒。

- 以 Hyperf、Swoole 建立高併發採集服務，RabbitMQ 調度與削峰
- 可配置式採集框架，規則存在 MongoDB，新站點改用設定接入
- Elasticsearch 全文搜尋，Canal 將 MySQL 即時同步到索引
- Redis 與 APCu 多層快取，降低資料庫負載

- 約 2 天 → 半天內: 新站點接入時間
- 約 2 秒 → 300 毫秒內: 搜尋回應時間
- 十萬級: 每日處理新聞資料
- 數千篇: 尖峰每分鐘採集量

搜尋時間比較的是全文搜尋上線前與上線後的回應，不是公開的百分位測試，也沒有留下查詢樣本數。

## 履歷問答助手

本站・規則工具＋只看履歷的模型

訪客會自由提問。模型若編造公司、數字或日期，履歷就不可信；模型逾時或失敗時，頁面也不能沒有答案。

- 頁面先把已知問題交給履歷工具：公司、技能、技術棧、學歷、證照、帶隊經歷與聯絡方式
- 對不上的問題交給只看精簡履歷的模型，並附上問題命中的那一段工作內容
- 模型沒有工具、沒有金鑰。問題最多 300 字，10 秒逾時，並限制來源網域與請求頻率
- 逾時、非成功回應或空回答時，改回規則答案

- 同一份履歷: 已知問題
- 只依履歷回答: 其餘問題
- 回到規則答案: 模型失敗時

就是本頁的問答助手，可以直接試。沒有另附評分基準。

## 主力技術

- AI Agents (核心): 設計任務導向的 agent 工作流程、工具呼叫，以及可靠的自動化路徑。
- Laravel (後端): 打造應用後端、管理流程、API 與服務整合。
- Golang (服務): 以 Go 寫內部打卡系統的後端，供 Flutter App 呼叫。系統只在內部使用，沒有公開的原始碼。
- Python (自動化): 做 AI 工作流程原型、資料工具、腳本與後端自動化。
- Node.js (執行環境): 串接 API、agent 工具、網路服務與非同步整合。

## 工作中用過的技術

### 語言與框架
- PHP: 約 10 年
- Hyperf: 約 6 年
- Swoole: 約 6 年
- Python: 約 2 年
- JavaScript: 約 3 年
- C#: 約 1 年
- ASP.NET: 約 1 年
- Delphi: 約 2 年
- ActiveX: 約 2 年
- Composer: 約 2 年

### 資料庫、快取與搜尋
- MySQL: 約 11.5 年
- MongoDB: 約 7.5 年
- Redis: 約 8.5 年
- APCu: 約 6 年
- Elasticsearch: 約 6 年
- Canal: 約 6 年
- SQL Server: 約 1 年

### 訊息與架構
- Kafka: 約 1.5 年
- RabbitMQ: 約 7 年
- Microservices: 約 1.5 年
- API Gateway: 約 1.5 年
- WebSocket: 約 1.5 年
- RESTful API: 約 2 年
- Third-party APIs: 約 2.5 年
- Account Ledger: 約 1.5 年
- AI: 約 1.5 年

### 基礎設施
- Linux: 約 2 年
- Nginx: 約 2 年
- CDN: 約 1.5 年
- S3: 約 2 年

## 學歷

- 資訊工程 學士, 中信科技大學 (2004–2006)
- 五年制專科 電腦應用工程, 中信科技大學 (1998–2003)

## 證照

- 電腦軟體應用 丙級技術士 (國家技術士證照)
- 電腦硬體裝修 丙級技術士 (國家技術士證照)
- SCJP 5.0（Sun Certified Java Programmer） (國際證照)

### Kos 有多少年的軟體開發經驗？
18 年以上。2008 年入行至今待過 7 間公司，從企業網站、晶片卡身分驗證，一路做到新聞搜尋、分佣帳務與支付錢包系統，其中兩段擔任技術主管。

### Kos 擅長哪些技術？
後端與 AI Agent 工作流程。語言以 PHP（Laravel、Hyperf、Swoole）、Golang、Python、Node.js 為主。Golang 這段是內部打卡系統的後端，供 Flutter App 使用，沒有公開的原始碼。資料層熟悉 MySQL、Redis、MongoDB、Elasticsearch，並以 Kafka、RabbitMQ 處理事件驅動與非同步任務。

### Kos 的 AI Agent 做了什麼？
本站的履歷問答助手。已知問題用履歷工具回答，涵蓋經歷、技能、學歷、證照與聯絡方式。其餘問題交給只看這份履歷的模型。問題最多 300 字，10 秒逾時；逾時、錯誤或空回答時，改回規則答案。模型沒有工具，也沒有金鑰。

### Kos 帶過多大的團隊？
最近擔任資訊主管（2024/11 – 2026/05），帶領 13 人以上的團隊，同時是支付、錢包與內容服務平台的技術負責人：規劃架構、分派工作、Code Review，並協調產品、客服、風控與財務等部門。

### Kos 做過哪些金流系統？
完整的資金鏈路：支付訂單、充值、提款、對帳，以及會員錢包與資金流水。以交易冪等與分散式鎖確保餘額一致，用補償、重試與自動對帳處理異常，並以 Kafka、Redis、MongoDB 建立事件驅動架構承接高併發交易。更早也重構過多層通路分佣與帳務結算系統。

### Kos 有哪些可量化的成果？
在傲創科技股份有限公司從零建置新聞採集與全文搜尋平台：每日穩定處理十萬級新聞資料、尖峰每分鐘數千篇，新站點接入時間由約 2 天縮短至半天內。搜尋回應時間是全文搜尋上線前約 2 秒、上線後 300 毫秒內，不是公開的百分位測試，也沒有留下查詢樣本數。

### Kos 的學歷與證照？
中信科技大學資訊工程學士；持有電腦軟體應用 丙級技術士、電腦硬體裝修 丙級技術士、SCJP 5.0（Sun Certified Java Programmer）。

### 如何聯絡 Kos？
透過本站的聯絡表單（/contact/），或 GitHub、LinkedIn 聯絡。Kos 正在尋找資深後端工程師或技術主管的機會，全職、約聘、接案皆可，台灣、海外或遠端都可以。

---

Focused on payment reliability, ledger consistency, and failure recovery. Former tech lead for payment, wallet, and content platforms, leading a 13+ person team and using idempotency, distributed locks, and automated reconciliation to prevent double charges and keep the books consistent. Now building AI agent workflows and backend services with Laravel, Golang, Python, and Node.js.

## Open to new roles
- Roles: Senior backend engineer, tech lead
- Type: Full-time, contract, freelance
- Where: Taiwan, overseas, or remote
- Languages: Chinese and English

## Prologue (2008–2013): Learning on client projects

Kos started in 2008 at ZhiShi Technology, running several corporate, campaign and custom sites at once. Then came ASP.NET and C# corporate sites and internal tools, and at YiTou Net, Delphi ActiveX controls wired to smart-card readers: ID verification and secure login for government and enterprise clients.

## Chapter 1 (2014–2018): First team, first ledger

At Happy Home Realty, Kos became engineering manager of a team of up to four, owning everything from architecture to Linux deployment. At ShengLian Network came a rebuild of the multi-tier commission and settlement system: a configurable rule engine, an account ledger and fund audits. From here on, the money had to add up.

## Chapter 2 (2019–2024): Scale and speed

At AoChuang Technology, Kos built a news crawling platform and full-text search from scratch, processing news at the 100K-per-day scale. A configurable crawling framework cut new-site onboarding from about 2 days to under half a day; search went from about 2 s to under 300 ms.

## Chapter 3 (2024–2026): Every payment counted once

As tech lead for payment, wallet and content platforms, Kos managed a team of 13+. Callback resends and network retries can process the same payment twice, so idempotency, distributed locks and automated reconciliation make each transaction post once and keep balances and ledgers matched.

## Now: AI Agents

Today Kos builds AI agent workflows and backend services with Laravel, Golang, Python and Node.js, and puts AI to work on the job: analyzing commission rules, detecting accounting anomalies, parsing documents, and assisting development and monitoring. The next chapter is open.

18 years in software across 7 companies and 9 key projects.

## 2024/11 – 2026/05 · Engineering Manager, Undisclosed company

- Led a team of 13+

Stack: Kafka, Redis, MongoDB, MySQL, API Gateway, Microservices, WebSocket, CDN

### Payment System
- Tech lead for payments; planned system architecture and technical design
- Planned team work, ran code reviews and mentored engineers
- Aligned product, support, risk and finance on scope and timelines
- Built payment orders, top-ups, withdrawals and reconciliation
- Built idempotency, compensation, retry and risk controls
- Introduced event-driven architecture with Kafka, Redis and MongoDB
- Tuned MySQL queries and high-concurrency transaction flows
- Set up transaction monitoring, alerting and automated ops

### Wallet System
- Tech lead for wallets; designed high-concurrency fund transactions
- Assigned work and tracked project progress
- Coordinated payments, finance, risk and QA on integration
- Built member wallets, accounting and fund ledgers
- Kept balances consistent with idempotency and distributed locks
- Used Redis caching to speed up high-concurrency transactions
- Built automated reconciliation, compensation and risk audits
- Designed a highly available wallet service

### Content Platform
- Led architecture and development planning
- Planned team work and coordinated third-party vendors
- Coordinated front end, back end, QA and ops to ship
- Built the public-facing and admin sides of the content platform, plus third-party content integrations
- Built an API gateway and microservices
- Added Redis, CDN and WebSocket for performance
- Optimized APIs, caching and high-traffic access
- Set up monitoring, alerting and AI-assisted development

## 2019/01 – 2024/10 · Senior PHP Engineer, AoChuang Technology

- Internet, 100–500 staff

Stack: PHP, Hyperf, Swoole, RabbitMQ, Redis, APCu, MongoDB, Elasticsearch, Canal, MySQL

- Site onboarding: ~2 days → < half a day
- Search latency: ~2 s → < 300 ms
- News processed daily: 100K-scale

### News Crawling & Full-Text Search
- Built a news crawling platform and full-text search from scratch
- Built high-concurrency crawlers with Hyperf and Swoole
- Used RabbitMQ for scheduling, load leveling and async jobs
- Built multi-level caching with Redis and APCu to cut DB load
- Stored crawl rules and unstructured data in MongoDB
- Built an Elasticsearch search platform with index management
- Synced MySQL to Elasticsearch in real time with Canal
- Built crawl-rule management, retries, error handling and monitoring
- Built a configurable crawling framework to onboard sites faster
- Set up logging, monitoring and alerting

Results
- Built a high-concurrency crawling and search platform from scratch
- Reliably processed news data at the 100K scale daily
- Cut new-site onboarding from about 2 days to under half a day
- Cut search response time from about 2 s to under 300 ms
- Built an extensible framework that sped up later work
- Built a highly available, scalable search architecture

## 2016/10 – 2018/01 · Senior R&D Engineer, ShengLian Network

- Systems integration, 1–30 staff
- Zhongzheng, Taipei

Stack: RabbitMQ, Redis, Account Ledger, AI

### Commission & Settlement
- Rebuilt the multi-tier channel commission and settlement system
- Built a configurable rule engine for multi-tier agents, channels and tiered rates
- Built commission calculation, batch settlement and accounting
- Used RabbitMQ for async batch settlement and scheduling
- Built an account ledger and fund auditing
- Added Redis caching to cut repeat computation
- Built transaction monitoring, compensation, retries and reconciliation
- Streamlined settlement and accounting for stability and scale
- Used AI to analyze rules, detect anomalies and improve accounting

Results
- Rebuilt the multi-tier commission and settlement system
- Shipped a rule engine for multi-channel, multi-tier and custom plans
- Raised settlement throughput with async batches on RabbitMQ
- Improved accounting consistency with a full ledger and audits
- Made commission rules configurable in the admin, reducing maintenance cost and speeding up delivery
- Built a highly available, scalable settlement architecture

## 2014/08 – 2016/10 · Engineering Manager, Happy Home Realty

- Led a team of up to 4
- Real estate, 100–500 staff
- West Central, Tainan

Stack: PHP, Python, MySQL, RESTful API, Composer, S3, Linux, Nginx

### Enterprise Management System
- As tech lead, planned the architecture, technology choices, and development process
- Assigned work, ran code reviews, and coordinated product, sales, QA, and other teams to deliver projects
- Led the core platform architecture and core modules
- Built internal systems and back-end services in PHP and Python
- Designed RESTful APIs for internal and external data exchange
- Designed MySQL schemas and indexes; tuned SQL
- Introduced Composer and a standard dev workflow
- Built PDF parsing, auto-import and data cleaning
- Built data collection, crawlers and sync services
- Designed file management on S3 object storage
- Built upload, download and permission APIs
- Set up Linux and Nginx, deployments and releases
- Set up monitoring, operations and incident handling
- Used AI to support document parsing, data analysis, and system performance optimization

Results
- Built a highly available platform and back-end architecture
- Integrated internal and external systems via APIs
- Automated PDF parsing and import, cutting manual work
- Automated data collection and sync for fresher data
- Built a central file management platform
- Standardized the dev process for speed and maintainability

## 2011/10 – 2013/10 · Engineer, YiTou Net

- Software services, 1–30 staff
- West Central, Tainan

Stack: PHP, Delphi, ActiveX, MySQL, JavaScript

### Smart-Card ID Verification
- Built and maintained systems for government and enterprise clients
- Ran client interviews, requirements analysis and planning
- Mapped client workflows and proposed technical solutions
- Coordinated scope, schedule and acceptance with PMs and clients
- Handled requirements analysis, system design, development, testing, deployment, and operations
- Built Delphi ActiveX controls to integrate smart-card readers
- Built certificate reading, ID verification and secure login
- Built the verification platform and back end in PHP
- Connected ActiveX to the PHP back end for login and members
- Built member data, permissions and login history
- Built RESTful APIs between front end, back end and other systems
- Solved browser, card-reader and back-end integration issues
- Handled testing, troubleshooting and technical docs
- Improved login flow, stability and security

Results
- Delivered systems for government and enterprise clients
- Built a smart-card ID verification and secure login platform
- Integrated Delphi ActiveX with smart-card readers
- Built PHP login, authorization, member and permission management
- Improved browser and card-reader compatibility
- Took clients through rollout and acceptance

## 2010/10 – 2011/10 · Engineer, XiaoBing LiDaGong

- Software services, 1–30 staff
- North, Tainan

Stack: ASP.NET, C#, SQL Server, JavaScript

### Corporate Sites & Internal Tools
- Built and maintained corporate sites and internal systems, from requirements to release
- Built ASP.NET / C# business logic: members, data entry, search and reports
- Integrated SQL Server: data access, query tuning and validation
- Handled testing, troubleshooting and version updates

## 2008/02 – 2010/10 · Engineer, ZhiShi Technology

- Internet, 30–100 staff
- West Central, Tainan

Stack: Third-party APIs

### Corporate & Campaign Sites
- Built corporate, campaign, and custom platforms across requirements, design, development, testing, deployment, and maintenance
- Planned features and data flows; built databases, back ends and templates
- Built and integrated third-party APIs and services
- Ran several projects in parallel and kept them on schedule
- Handled fixes, optimization, new features and support

Results
- Delivered corporate, campaign and admin sites: members, surveys and more
- Built reusable modules that shortened delivery
- Integrated many third-party services, cutting maintenance cost
- Kept quality and deadlines across parallel projects

## Payment and wallet systems

Tech lead, 13+ person team, 2024–2026

Under high concurrency, a resent callback or a network retry can process the same payment twice, and member balances must always match the ledger.

- Payment idempotency, distributed locks, and balance consistency checks, so each transaction posts once
- Compensation and retry for failed transactions, with risk-control audits
- Event-driven architecture on Kafka, Redis, and MongoDB
- Automated reconciliation, plus transaction monitoring and alerts
- Verified with tests for resent callbacks, concurrent requests, and compensation

- No double charges: Idempotency and locks
- Auto reconciliation: Compensation and audits

The company name and business figures are not public. What can be stated is the test coverage: resent callbacks, concurrent requests, and compensation. No live transaction volume is published.

## News crawling and full-text search

AoChuang Technology, senior engineer, 2019–2024

Every new news site needed its own crawler, about 2 days of work, and full-text search took about 2 seconds.

- High-concurrency crawling on Hyperf and Swoole, scheduled and buffered with RabbitMQ
- A configurable crawling framework with rules in MongoDB, so new sites are added by configuration
- Elasticsearch full-text search, kept in sync with MySQL in real time through Canal
- Multi-level caching with Redis and APCu to take load off the database

- ~2 days → under half a day: New site onboarding
- ~2 s → under 300 ms: Search response time
- 100K+: News items processed daily
- Thousands: Items per minute at peak

The search time compares the response before and after full-text search went live. It is not a published percentile, and the query sample size was not kept.

## Resume Q&A assistant

This site · resume tools plus a resume-only model

Visitors ask free-form questions. A model that invents employers, numbers, or dates would make the resume untrustworthy, and a model failure should still leave an answer on the page.

- The page sends known questions to resume tools: company, skill, stack, education, certifications, leadership, and contact
- Other questions go to a model that sees a compact resume plus the bullets of the one job the question matches
- The model has no tools and no secrets. Questions stop at 300 characters, the call times out at 10 seconds, and requests are limited by origin and rate
- A timeout, a failed response, or an empty answer falls back to the rule-based reply

- Same resume data: Known questions
- Resume only: Other questions
- Rule fallback: When the model fails

This is the assistant on this page. There is no separate score.

## Core skills

- AI Agents (Focus): Designing task-oriented agent workflows, tool use, and reliable automation paths.
- Laravel (Backend): Building application backends, admin workflows, APIs, and service integrations.
- Golang (Services): Go backend for an internal clock-in system, called by a Flutter app. Internal only, with no public repository.
- Python (Automation): Prototyping AI workflows, data utilities, scripts, and backend automation.
- Node.js (Runtime): Connecting APIs, agent tools, web services, and asynchronous integrations.

## Used at work

### Languages & frameworks
- PHP: ~10 yrs
- Hyperf: ~6 yrs
- Swoole: ~6 yrs
- Python: ~2 yrs
- JavaScript: ~3 yrs
- C#: ~1 yrs
- ASP.NET: ~1 yrs
- Delphi: ~2 yrs
- ActiveX: ~2 yrs
- Composer: ~2 yrs

### Data, cache & search
- MySQL: ~11.5 yrs
- MongoDB: ~7.5 yrs
- Redis: ~8.5 yrs
- APCu: ~6 yrs
- Elasticsearch: ~6 yrs
- Canal: ~6 yrs
- SQL Server: ~1 yrs

### Messaging & architecture
- Kafka: ~1.5 yrs
- RabbitMQ: ~7 yrs
- Microservices: ~1.5 yrs
- API Gateway: ~1.5 yrs
- WebSocket: ~1.5 yrs
- RESTful API: ~2 yrs
- Third-party APIs: ~2.5 yrs
- Account Ledger: ~1.5 yrs
- AI: ~1.5 yrs

### Infrastructure
- Linux: ~2 yrs
- Nginx: ~2 yrs
- CDN: ~1.5 yrs
- S3: ~2 yrs

## Education

- B.S., Computer Science & Information Engineering, Chung Hsin University of Technology (2004–2006)
- 5-year college, Computer Application Engineering, Chung Hsin University of Technology (1998–2003)

## Certifications

- Computer Software Application, Level C Technician (Taiwan national certification)
- Computer Hardware Repair, Level C Technician (Taiwan national certification)
- SCJP 5.0 (Sun Certified Java Programmer) (International certification)

### How much experience does Kos have?
18+ years, since 2008, across 7 companies. The work grew from company websites and smart-card login systems into news search, commission settlement, and payment and wallet platforms. Kos has led engineering teams twice.

### What technologies does Kos work with?
Kos builds backend systems and AI agent workflows with PHP (Laravel, Hyperf, Swoole), Golang, Python and Node.js. Golang here is the backend of an internal clock-in system used by a Flutter app; that system has no public repository. On the data side Kos works with MySQL, Redis, MongoDB and Elasticsearch, and uses Kafka and RabbitMQ for event-driven and asynchronous processing.

### What AI agent work has Kos shipped?
The Q&A assistant on this site. Known questions use resume tools for jobs, skills, education, certifications, and contact details. Other questions go to a model that answers only from that same resume. Questions stop at 300 characters and the call times out at 10 seconds. A timeout, an error, or an empty answer falls back to the rule-based reply. The model has no tools and no secrets.

### Has Kos led a team?
Yes. As Engineering Manager (2024/11 – 2026/05) Kos led a team of 13+ and was the technical lead for the payment, wallet and content platforms: designing the architecture, assigning work, reviewing code, and coordinating product, support, risk and finance.

### What payment systems has Kos built?
The full money flow: payment orders, top-ups, withdrawals and reconciliation, plus member wallets and transaction ledgers. Balances stay correct through idempotent transactions and distributed locks; failures are handled with compensation, retries and automated reconciliation; Kafka, Redis and MongoDB carry the high-concurrency event flow. Earlier, Kos rebuilt a multi-tier commission and settlement system.

### What measurable results has Kos delivered?
At AoChuang Technology, Kos built a news crawling and full-text search platform from scratch. It processes news items at the 100K scale every day, thousands a minute at peak, and cut new-site onboarding from about 2 days to under half a day. Search response time went from about 2 s before full-text search went live to under 300 ms after. That is not a published percentile, and the query sample size was not kept.

### What education and certifications does Kos have?
B.S., Computer Science & Information Engineering, Chung Hsin University of Technology. Certifications: Computer Software Application, Level C Technician; Computer Hardware Repair, Level C Technician; SCJP 5.0 (Sun Certified Java Programmer).

### How can I contact Kos?
Use the contact form at /contact/, or reach Kos on GitHub or LinkedIn. Kos is open to senior backend engineer and tech lead roles: full-time, contract or freelance, in Taiwan, overseas or remote.