Guo Qiang · Product Engineer

Projects
& Labs

以“AI 辅助编程 + 干中学”落地的确定性开源系统。涵盖 AI 工作流编排、DAG 状态机、受限解码与硬件实验。 Deterministic open-source systems built via AI Pair Programming & Learning by Doing. Covering AI workflow orchestration, DAG state machines, constrained decoding, and hardware experiments.

FLAGSHIP · CORE PROJECT

PatchCat

AI 提示流编排器与 DAG 状态机 AI Prompt Flow Orchestrator & DAG State Machine

DAG State Machine Kahn's Algorithm Constrained Decoding Zod Guard Local-First

纯前端、本地优先运行的确定性 AI 工作流编排系统。通过 Kahn 拓扑排序调度避免环路死锁，内置 L1 Logits 掩码受限解码与 L2 Zod 契约双层防线拦截脏输出，提供 5 秒 Web Worker 脚本沙箱与死循环看门狗熔断。

A client-only, local-first deterministic AI workflow orchestrator. Employs Kahn's topological sort for acyclic graph scheduling, two-tier constrained decoding (Logits masking + Zod contract), and Web Worker watchdog sandbox to circuit-break infinite loops.

在线体验 LIVE DEMO ↗ GITHUB REPO ↗ 架构与实验矩阵 ↗ ARCHITECTURE MATRIX ↗

Creative Labs & Hardware

查看所有项目与实验工坊 VIEW ALL PROJECTS & LABS ↗

ONLINE ↗

玄学工具集 Metaphysics Tools ↗

八字 / 六爻 / 紫微 / 周易算法建模 Bazi / I Ching / Ziwei Algorithm Engine

GITHUB ↗

Kindle 墨水屏看板 Kindle Weather Station ↗

电子墨水屏 / 7段数码管 / 桌面时钟 E-Ink Display / 7-Segment Clock / Station

Writings

查看所有文章归档 VIEW ALL POSTS ↗

Product Genesis & Real-World Demand 2026-10-22

为什么我开始做 FerryCat：当你的电脑上同时开着两个 AI 编程软件 Why I Started Building FerryCat: When Two AI Coding Hosts Live on the Same Desktop

在借助 AI 辅助编程推进日常开发时，让同一个模型既写代码又做代码审查，本质上是“自己批改自己的作业”。为了打破同源盲区，我开始在电脑上同时打开多个不同 AI 编程宿主跨窗口互审，却立刻被“人肉剪贴板搬运”、“未落盘旧代码盲审”和“两千字散文式回复”折磨得苦不堪言。本文不谈底层代码细节，只聊 FerryCat 诞生的真实需求起源、早期 Demo 阶段向物理限制低头的诚实取舍、上线第一天帮我抓出 9 个真实漏洞的实战记录，以及迈向跨宿主全自动对话的终极产品形态。 Relying on the same LLM to both write and audit code is equivalent to grading one's own homework. Opening multiple desktop AI coding hosts to cross-review code solves model blind spots, but immediately traps the developer in clipboard copy-paste fatigue, silent blind reviews on unsaved stale files, and unstructured essay responses. Setting aside low-level code details, this post explains the real-world demand behind FerryCat, its honest Demo-stage trade-offs under desktop process constraints, how its very first self-audit across three heterogeneous models caught 9 real production bugs, and its ultimate roadmap toward zero-intervention cross-agent dialogue.

9 MIN READ READ ARTICLE ↗

Organizational Engineering & System Boundaries 2026-10-20

做东西变快了，钱却更难挣了：AI 正在抽空组织的腰部骨干 Building Faster, Earning Harder: How AI Hollows Out the Organizational Backbone

Anthropic 内部数据显示，AI 代理工时已达人类员工 6 倍，但初级技术岗位招聘比例暴跌至 9%。本文从 Product Engineer 视角深度解构这场组织形变：初级练习场瓦解，中层自保推卸责任，导致资深老兵被系统性架空。当低成本代码撞上公域商业化暗礁与跨系统接口无日志暗病，生产力过剩的终局，最稀缺的永远是能为商业转化与系统边界兜底负责的人。 Internal data reveals AI agents working 6x human hours, yet junior engineering hiring plummeted to 9%. From a Product Engineer's perspective, this essay dissects the organizational hollow-out: entry-level practice grounds are wiped out while middle-layer self-preservation isolates senior talent. As zero-cost code generation crashes against real-world traffic dark reefs and cross-system silent interface failures, organizations will confront the ultimate truth: in an era of production abundance, genuine value belongs solely to those who sign off on commercial viability and boundary defense.

10 MIN READ READ ARTICLE ↗

Browser-Native Hybrid Search & RRF 2026-10-18

Zero-Backend AI: Running Local Embeddings, BM25, and Hybrid RRF Search 100% in the Browser Zero-Backend AI: Running Local Embeddings, BM25, and Hybrid RRF Search 100% in the Browser

Why pay monthly cloud bills for hosted vector databases when modern browsers can run quantized embeddings, an in-memory BM25 index, and reciprocal rank fusion locally? Here is how we engineered a zero-backend hybrid search pipeline in PatchCat with 0 server dependencies, sub-50ms latency, and 100% client-side privacy. Why pay monthly cloud bills for hosted vector databases when modern browsers can run quantized embeddings, an in-memory BM25 index, and reciprocal rank fusion locally? Here is how we engineered a zero-backend hybrid search pipeline in PatchCat with 0 server dependencies, sub-50ms latency, and 100% client-side privacy.

6 MIN READ READ ARTICLE ↗

Season 1 Finale & Architecture Retrospective 2026-10-18

PatchCat 第一季收官复盘：从零构建纯端侧 AI 工作流引擎的架构全景与“边写边学”实战手记 PatchCat Season 1 Finale: Architectural Blueprint and Learning-by-Doing Retrospective of a Browser-Native AI Workflow Orchestrator

历经 26 篇持续攻坚与 592 项全量自动化测试，PatchCat 在 0.4.x 阶段完成了纯浏览器端 Local-First AI 工作流引擎的全部底层原语构建。本文作为开源第一季的收官复盘，系统梳理确定性 DAG 状态机调度、端侧向量与混合检索、Web Crypto 硬件级安全脱敏与版本快照 Diff 架构全景；深入复盘人机双向共创与“像在实验室做功课般边写边学”的研发心流，坦承端侧算力边界与架构代价，正式宣布核心原语封箱并迈向 v0.5.0 稳态新阶段。 Spanning 26 technical notes and 592 automated tests, PatchCat has completed all browser-native Local-First primitives throughout its 0.4.x era. As the Season 1 finale, this retrospective outlines the full architecture: deterministic DAG state machine scheduling, client-side vector embeddings, Web Crypto vaults, and structural diffing; documents human-AI co-creation and the lab-like learning-by-doing workflow; details architectural trade-offs; and seals core primitives as we step into v0.5.0.

20 MIN READ READ ARTICLE ↗

Work

深耕企业级 SaaS、大流量高并发架构与 AI 原生系统工程。 Covering enterprise SaaS, high-concurrency systems, and AI-native engineering.

2024.09 – PRESENT

喵嗷呜科技工作室 Miaoawu Tech Studio

联合创始人 / 技术产品负责人 Co-founder / Tech Product Lead

主导企业级 AI Agentic Workflow 与自动化工具链搭建，基于 LLM API、n8n 构建自动化流水线，设计跨平台数据同步与统一 API 接口。 Architected enterprise AI automation pipelines & agentic workflows (LLM APIs, n8n), unified cross-platform API contracts, and managed rapid agile prototypes.

2022.02 – 2024.09

极狐 GitLab JiHu (GitLab)

产品专家 / SaaS 与平台产品负责人 Senior Product Manager / Product Specialist

主导中国区 SaaS 平台与本地化企业版产品规划，管理跨时区国际化分布式产研协同，推动路线图制定与高可用企业级功能端到端交付。 Owned product planning and quarterly roadmaps for China SaaS & Self-Managed enterprise editions, coordinating multi-timezone international R&D teams.

2019.08 – 2022.02

上海仙泽分析仪器 Shanghai Xianze Analytical Instruments

软件与平台产品负责人 Software & Platform Product Lead

面向科研与工业实验室场景，主导高精度光谱分析工作站软件与 SaaS 数据平台的产品架构与交付，搭建 Jira 敏捷管理体系。 Led product architecture and delivery for optical workstation software and laboratory SaaS data platforms, standardizing Jira-based agile workflows.

2018.06 – 2019.05

上海钦文网络科技 Shanghai Qinwen Network Technology

高级产品经理 Senior Product Manager

通过全流程 A/B 测试优化用户注册与付费转化链路，结合精细化用户分层运营策略，推动注册转化率提升 30%+，DAU 提升 20%+。 Designed end-to-end A/B tests to optimize registration and onboarding funnels, boosting conversion rate by 30%+ and DAU by 20%+.

2016.09 – 2018.03

哔哩哔哩 (Bilibili) Bilibili

高级产品经理 Senior Product Manager

负责移动主站跨端组件与 Hybrid 架构升级，确保大流量峰值下的系统流畅度与交付一致性；搭建全栈转化漏斗与埋点监控指标体系。 Deployed cross-platform UI frameworks and Hybrid infrastructures for peak traffic; established full-stack event tracking and funnel monitoring metrics.

2014.03 – 2016.03

携程 (Ctrip) Ctrip

产品经理 (营销与技术平台) Product Manager (Technology PM / Marketing)

主导 APP 深度链接（Deep Linking）与 URI 路由标准化系统重构，使全站订单转化率提升 70%~80%；规划原生广告全生命周期看板与标准化营销 API。 Spearheaded APP Deep Linking & URI routing overhaul, boosting order conversion rate by 70%~80%; designed marketing automation APIs and real-time dashboards.

2012.07 – 2014.01

北京联银通科技 Beijing Lianyintong Technology

软件工程师 Software Engineer

负责银行核心数据聚合平台（DCP）的后端开发、接口编写、自动化测试与生产环境部署，积累扎实的底层软件架构与数据处理经验。 Developed backend data aggregation modules and automated test pipelines for core banking platforms, building a solid software architecture foundation.

About
The Process

作为一名 Product Engineer，我不相信空谈愿景或纯靠 Prompt 抽卡能做出真正可靠的生产级系统。

大模型天然具备概率与不确定性，但企业的业务逻辑和自动化流水线必须是确定性的。我的工程目标，就是用 DAG 状态机、Kahn 拓扑排序和受限解码 等确定性铁轨，把不可控的概率收敛在安全边界之内。

在技术实现上，我坚持“平台底蕴 + 商业增长”的双轨实践：既要把控底层系统的稳定与防死锁，又要站在真实用户视角把上手门槛做低。

As a Product Engineer, I don't believe vague vision statements or prompt tinkering alone produce reliable production systems.

LLMs are inherently probabilistic, but real business automation requires determinism. My engineering focus is laying white-box rails—DAG state machines, Kahn topological scheduling, and constrained decoding—to keep non-deterministic models bounded within safe contracts.

In practice, I balance platform engineering rigor with product growth: maintaining rock-solid deadlock prevention beneath the hood while obsessing over zero-friction user onboarding.

Projects& Labs