Nghĩ như PM, làm như builder. Tư vấn, xây dựng và kiểm định sản phẩm AI cho doanh nghiệp, dựa trên những sản phẩm đã chạy thật.https://vuihoc.ai/https://vuihoc.ai/insights/5-cloud-llm-biases-5-levers/https://vuihoc.ai/insights/5-cloud-llm-biases-5-levers/Why Claude / GPT / Gemini get it wrong so often when writing workflow docs, PRDs, BRDs — and why you can't 'fix' the weights. You can only build guardrails around them.Thu, 05 Dec 2024 00:00:00 GMThttps://vuihoc.ai/insights/7-pm-decisions-building-eval-framework/https://vuihoc.ai/insights/7-pm-decisions-building-eval-framework/Field notes through a PM lens: building Eval-Framework for an SMB use case (~15k audits/mo). The AI engineering pair proposed 7 decisions that could have burned budget, leaked compliance, and shipped the wrong roadmap. PM catches + reframes. Counter-prompt playbook below.Tue, 24 Feb 2026 00:00:00 GMThttps://vuihoc.ai/insights/adlc-app-development-lifecycle-for-solo-builders/https://vuihoc.ai/insights/adlc-app-development-lifecycle-for-solo-builders/Building alone with an AI coding agent removes every natural checkpoint a team gives you for free — no PM to catch scope creep, no reviewer to catch a dropped requirement. ADLC is the lifecycle apparatus I built to put those checkpoints back: ten phases, a gate discipline, and an autonomy contract that says exactly when the agent stops and when it doesn't.Sat, 11 Jul 2026 00:00:00 GMThttps://vuihoc.ai/insights/ai-canon-crawler-vendor-doc-rag/https://vuihoc.ai/insights/ai-canon-crawler-vendor-doc-rag/Personal-RAG kept hallucinating vendor specs because Marc's notes and Anthropic's docs lived in the same index. The fix: a dedicated canon workspace, a crawler that only ingests authoritative sources, and a routing rule that prefers canon when the question is about pricing, model size, or API behavior.Sun, 24 May 2026 00:00:00 GMThttps://vuihoc.ai/insights/an-ai-that-audits-my-ai-infrastructure/https://vuihoc.ai/insights/an-ai-that-audits-my-ai-infrastructure/A week spent hardening fukuro — a personal auditor for AI infrastructure — turned into a lesson in recursion. I watched the auditor turn on itself, and its most useful output was a list of the boring, quiet mistakes that both I and the AI helping me had missed.Sat, 04 Jul 2026 00:00:00 GMThttps://vuihoc.ai/insights/bookts-ios-app-built-100-percent-through-adlc/https://vuihoc.ai/insights/bookts-ios-app-built-100-percent-through-adlc/I built a 9-phase Agentic Development Life Cycle (ADLC) as reusable execution scaffolding, then ran a real App Store product — BookTS, an on-device EPUB-to-audiobook app — through all of it. Two of the phases caught bugs I would have shipped blind.Sat, 18 Jul 2026 00:00:00 GMThttps://vuihoc.ai/insights/building-adlc-in-two-weeks/https://vuihoc.ai/insights/building-adlc-in-two-weeks/I paused a real App Store build to spend two weeks building the pipeline that would run it instead — a ten-phase Agentic Development Life Cycle, built by the same AI coding agent that would later execute it. What got reused versus built, the planning audit I overrode to own a review engine instead of just calling one, and the morning that engine caught a bug in its own predecessor's verdict.Sun, 12 Jul 2026 00:00:00 GMThttps://vuihoc.ai/insights/competitor-research-30-parallel-ai-agents/https://vuihoc.ai/insights/competitor-research-30-parallel-ai-agents/Before building 'Jira for AI agents,' I ran one Claude prompt that fanned out into 111 sub-agent runs — roughly twenty minutes and 3.6 million tokens of work — across five research angles searched in parallel, 29 sources pulled, 133 claims extracted, then three independent skeptics voted on every claim. Only 9 survived. The verdict killed the idea in minutes. A PM's field notes on using AI to disprove yourself, cheaply.Wed, 03 Jun 2026 00:00:00 GMThttps://vuihoc.ai/insights/diagram-engine-text-to-diagram-for-ai-agents/https://vuihoc.ai/insights/diagram-engine-text-to-diagram-for-ai-agents/Why Mermaid/PlantUML break down when an LLM is the author. JTBD, syntax design constraints, and 37 layout rules that made the output good enough for a public playground.Mon, 08 Dec 2025 00:00:00 GMThttps://vuihoc.ai/insights/distilling-a-rag-knowledge-base/https://vuihoc.ai/insights/distilling-a-rag-knowledge-base/My personal RAG had retrieval near its ceiling but was still giving truncated, sometimes-hallucinated answers. So I stopped tuning the retriever and rewrote the corpus itself: raw transcripts distilled into atomic, dated, self-contained claims. Answer quality went up with statistical significance on two separate corpora, and the fix removed hallucination instead of adding it.Mon, 27 Jul 2026 00:00:00 GMThttps://vuihoc.ai/insights/eval-framework-pm-journey/https://vuihoc.ai/insights/eval-framework-pm-journey/Building a personal eval framework: from a 20-case smoke test → 192 stratified cases → a bake-off across 6 models. When 'guessing' which LLM is best gets expensive and wrong.Wed, 21 Jan 2026 00:00:00 GMThttps://vuihoc.ai/insights/governing-a-claude-code-rollout-kame/https://vuihoc.ai/insights/governing-a-claude-code-rollout-kame/An engineering org rolling out AI coding agents asked for seven controls: token visibility per developer, daily caps, spend alerts, a central off switch, a routing guard, CI failure cost and per-developer KPIs. What the vendor console covers, what it does not, and the small control plane I built to close the gap, with the numbers it produced.Sat, 12 Sep 2026 00:00:00 GMThttps://vuihoc.ai/insights/harness-engineering-scaffolding-is-the-product/https://vuihoc.ai/insights/harness-engineering-scaffolding-is-the-product/I ran two AI coding agents — Claude Code and Antigravity — through 100 turns of pair-programming over a Redis Streams channel I built, while they wrote real code against a live Redis. The code wasn't the lesson. What kept them working together was the harness. Field notes on engineering the scaffolding around an LLM.Thu, 28 May 2026 00:00:00 GMThttps://vuihoc.ai/insights/health-coach-ai-nudge-sleep-adhd/https://vuihoc.ai/insights/health-coach-ai-nudge-sleep-adhd/I've worn a Garmin for 2 years — sleep, HRV, RHR data all there but zero action. This post logs my build through a PM lens: JTBD, existing alternatives, MVP scope decisions, ADHD-first design principles, and a V2 roadmap for a personalized AI sleep coach over Telegram.Sat, 09 Aug 2025 00:00:00 GMThttps://vuihoc.ai/insights/idea-to-ship-pipeline-generator-and-judges/https://vuihoc.ai/insights/idea-to-ship-pipeline-generator-and-judges/Five AI agents wired into one idea-to-ship lane — one generator, four judges, and a runtime they share. The architecture, an end-to-end walkthrough, and the honest list of what it can't do.Thu, 18 Jun 2026 00:00:00 GMThttps://vuihoc.ai/insights/in-app-purchases-app-store-connect-gotchas/https://vuihoc.ai/insights/in-app-purchases-app-store-connect-gotchas/I set up In-App Purchases for a real App Store app for the first time and burned about 20 hours on failures that had nothing to do with my code. Five concrete traps — an agreement that looks signed but isn't, a save button that silently discards your work, staggered sandbox propagation, a subscription you can't test because you already own the product, and an icon rejected for one invisible pixel property.Sun, 19 Jul 2026 00:00:00 GMThttps://vuihoc.ai/insights/knowledge-audit-3-layer-catch-ai-memory-lying/https://vuihoc.ai/insights/knowledge-audit-3-layer-catch-ai-memory-lying/Claude Code, Cursor, ChatGPT all ship 'persistent memory'. Six months in, the AI confidently asserts facts that are no longer true. I built a 3-layer audit plus a 4-tier cron to verify every claim.Sat, 08 Nov 2025 00:00:00 GMThttps://vuihoc.ai/insights/knowledge-audit-rag-kb-grok-99-percent/https://vuihoc.ai/insights/knowledge-audit-rag-kb-grok-99-percent/A real production use case: build a cross-source contradiction detector on Personal-RAG. Verify ground truth with Eval-Framework, swap Haiku → Grok 4.3, real accuracy 80% → 99%.Thu, 23 Apr 2026 00:00:00 GMThttps://vuihoc.ai/insights/llm-bake-off-6-judges-methodology/https://vuihoc.ai/insights/llm-bake-off-6-judges-methodology/Tested Claude (Haiku/Sonnet/Opus) vs Grok 4.3 vs GPT-5.4-mini vs Gemini Flash. N=30 dev + N=99 holdout. Stratified eval. Pairwise Cohen's kappa. Surprising winner.Sun, 22 Mar 2026 00:00:00 GMThttps://vuihoc.ai/insights/llm-os-and-second-brain/https://vuihoc.ai/insights/llm-os-and-second-brain/Interactive page visualizing Andrej Karpathy's philosophy for building an LLM Operating System and RAG architecture.Tue, 20 Aug 2024 00:00:00 GMThttps://vuihoc.ai/insights/local-vs-cloud-llm-voice-agent/https://vuihoc.ai/insights/local-vs-cloud-llm-voice-agent/I built an AI voice assistant for daily use to truly understand the local vs cloud LLM trade-off. The surprise lesson: for agentic products, $0.001/turn Cloud is cheaper than 'free' local — once you count true TCO. A decision framework for product teams and CTOs.Tue, 18 Feb 2025 00:00:00 GMThttps://vuihoc.ai/insights/mac-translator-force-click-menubar/https://vuihoc.ai/insights/mac-translator-force-click-menubar/Building a menubar translator triggered by Force Click anywhere on macOS. Why the input gesture matters more than the model, and the AX/permission rebuild trap that cost me two evenings.Fri, 10 Apr 2026 00:00:00 GMThttps://vuihoc.ai/insights/mail-assistant-adhd-inbox-triage/https://vuihoc.ai/insights/mail-assistant-adhd-inbox-triage/Field note: building a personal email triage agent that surfaces only what actually needs action. 568 messages → 7 P0/P1 rows. The 5 product decisions that killed 87% of the noise.Mon, 15 Sep 2025 00:00:00 GMThttps://vuihoc.ai/insights/notebooklm-podcast-prompt/https://vuihoc.ai/insights/notebooklm-podcast-prompt/Turn any document into a deep, vivid 15-minute Podcast Audio Overview entirely in Vietnamese with this prompt.Thu, 10 Oct 2024 00:00:00 GMThttps://vuihoc.ai/insights/personal-rag-kb-4-iterations/https://vuihoc.ai/insights/personal-rag-kb-4-iterations/Why I moved from Oracle ADB cloud → pgvector replica → local Mac — and what I learned measuring latency, quality, and cost at each version.Thu, 22 May 2025 00:00:00 GMThttps://vuihoc.ai/insights/prompt-engineering-101/https://vuihoc.ai/insights/prompt-engineering-101/A roundup of fundamental and advanced Prompt Engineering techniques for interacting effectively with LLMs.Sat, 15 Jun 2024 00:00:00 GMThttps://vuihoc.ai/insights/train-object-detection-realtime-web/https://vuihoc.ai/insights/train-object-detection-realtime-web/A PM journey: from 'the off-the-shelf API counts wrong' → fine-tuning YOLO → real-time on-device inference with WebGPU. Product decisions for building computer vision, not a code tutorial.Thu, 04 Jun 2026 00:00:00 GMThttps://vuihoc.ai/insights/welcome-to-vuihoc-ai/https://vuihoc.ai/insights/welcome-to-vuihoc-ai/First post — Introducing Vuihoc.AI and why I started journaling my AI research journey.Mon, 01 Apr 2024 00:00:00 GMThttps://vuihoc.ai/insights/work-channels-as-evidence-index/https://vuihoc.ai/insights/work-channels-as-evidence-index/Every mainstream productivity framework is team-level and warns you off individual evaluation. Managers still have to do it. A PM approach to reading Git, Jira, Confluence and Slack as evidence about a person, plus the design rules that keep it from turning into surveillance.Sun, 30 Aug 2026 00:00:00 GMT