{"slug": "why-chatgpt-fails-as-an-interview-tutor-and-how-i-built-a-better-one-with-claude", "title": "Why ChatGPT Fails as an Interview Tutor (And How I Built a Better One with Claude Code)", "summary": "A developer built an AI interview tutor using Claude Code's rule files and persistent memory system to overcome ChatGPT's tendency to be an \"attentive chat companion\" rather than a strict teacher. The system enforces hard-coded rules for mixed question sets, mandatory explanations for wrong answers, and spaced repetition scheduling, while also generating follow-up questions from the user's own project code. The open-source project, available on GitHub, combines cross-session memory with understanding—a capability the developer says no other tool offers.", "body_md": "When you use ChatGPT / Claude to prepare for interviews, have you ever encountered this situation:\n\nThe root cause is simple: **ChatGPT is an attentive chat companion, not a strict teacher.**\n\nIts working principle is \"make the user satisfied,\" not \"make the user learn.\" It thinks everything you say is great because it's afraid you'll be unhappy.\n\nYes. I did it using **Claude Code** (Anthropic's CLI coding tool).\n\nClaude Code has a unique capability: **Rule files (CLAUDE.md) + Persistent memory system**. In simple terms, you can \"program\" the AI's behavior by writing rule files, and these rules are automatically loaded in every conversation.\n\nI applied this capability to interview preparation.\n\nI wrote a set of configuration files, with two core components:\n\n```\nRules:\n1. Mixed question sets (3-4 modules, not just one direction)\n2. Correct answer → Brief confirmation + Related knowledge points → Next question\n3. Wrong answer → Must explain (including reference URLs) → Verification question → New question only after passing\n4. No skipping steps. Cannot ask the next question immediately after an answer.\n```\n\nThe difference from a regular AI: **The rules are hard-coded, not decided by the AI itself.** It can't be lazy and say \"Great,\" because the rule file tells it it must complete the explanation → verification process.\n\n| Knowledge Point | First Learned | D+1 | D+2 | D+4 | D+7 | D+15 | Status |\n|---|---|---|---|---|---|---|---|\n| Smart Pointers | 5/29 | ✅ | ✅ | Mastered | |||\n| Virtual Function Table | 5/29 | ❌ | Weak |\n\nAt the start of each session, Claude automatically checks which knowledge points are due for review and prioritizes review questions. **You don't need to remember when to review yourself.**\n\nSay \"That's it for today\" to trigger the wrap-up process:\n\nThis is the **AI version of the Feynman Technique** — not just feeling like you understand, but being able to explain it, with AI as the judge.\n\n| Capability | ChatGPT Interview | Anki | NotebookLM | SaaS Platforms | This Project |\n|---|---|---|---|---|---|\n| Cross-session Memory | None | Deck-level only | None | Partial | Yes |\n| Spaced Repetition | None | Yes | None | None | Yes |\n| Wrong → Explain → Verify | None | No (Only gives answer) | None | Partial | Yes |\n| Native Question Generation from Project Code | Need to manually paste code | N/A | Requires file upload | Not supported | Native Support |\n| Additional Cost | None | None | None | \\$29-300/time | No Extra Cost |\n\nThe core difference: **Other tools either have memory without understanding (Anki), or understanding without memory (ChatGPT), or neither (SaaS).** Claude Code's rule file + memory system is the only solution that combines both.\n\nAll interview tools on the market are \"I ask, you answer.\"\n\nThis project is different: **Asking follow-up questions starting from your own code.**\n\nFor example, my project has a `BoundedQueue`\n\n(Bounded Queue) using two `condition_variable`\n\n. During interview prep, Claude doesn't ask you \"Please explain condition_variable,\" but rather:\n\n\"Your BoundedQueue uses two condition_variables (\n\n`not_full_`\n\nand`not_empty_`\n\n). Can you use just one?\"\n\nThis kind of question forces you to **think about the principles from code you've written yourself**, giving the knowledge an anchor point, not just memorizing standard answers.\n\n```\ngh repo fork happiness-cheng/ai-interview-engine --clone\ncd ai-interview-engine\nclaude\n```\n\nI want to prepare for a C++ backend developer interview, aiming for a big tech internship. Let's start reviewing.\n\nThat's it. Claude will read the rule files and start the first round of questions.\n\nThe repository comes with an example question bank for C++ backend. If you use Java / Go / Frontend, replace it with the blank template:\n\n```\nknowledge/TEMPLATE.md → Copy to interview_tracker.md → Fill in your knowledge points\n```\n\nI'm a sophomore student preparing for the 2026 fall internship recruitment. Using this system for C++ backend interview prep, the most noticeable feelings are:\n\nThe underlying engine is general-purpose: **Question → Answer → Judge → Explain → Verify → Forgetting Curve Scheduling**.\n\nInterviews are just one scenario. You can use it to prepare for:\n\nGitHub: [https://github.com/happiness-cheng/ai-interview-engine](https://github.com/happiness-cheng/ai-interview-engine)\n\nIf it helps you, give it a star. If you have suggestions for improvement, feel free to open an issue.", "url": "https://wpnews.pro/news/why-chatgpt-fails-as-an-interview-tutor-and-how-i-built-a-better-one-with-claude", "canonical_source": "https://dev.to/guanyi_liu_21a5d7417eb332/why-chatgpt-fails-as-an-interview-tutor-and-how-i-built-a-better-one-with-claude-code-p94", "published_at": "2026-05-30 13:16:19+00:00", "updated_at": "2026-05-30 13:52:40.068251+00:00", "lang": "en", "topics": ["large-language-models", "ai-tools", "ai-products"], "entities": ["ChatGPT", "Claude", "Anthropic", "Claude Code"], "alternates": {"html": "https://wpnews.pro/news/why-chatgpt-fails-as-an-interview-tutor-and-how-i-built-a-better-one-with-claude", "markdown": "https://wpnews.pro/news/why-chatgpt-fails-as-an-interview-tutor-and-how-i-built-a-better-one-with-claude.md", "text": "https://wpnews.pro/news/why-chatgpt-fails-as-an-interview-tutor-and-how-i-built-a-better-one-with-claude.txt", "jsonld": "https://wpnews.pro/news/why-chatgpt-fails-as-an-interview-tutor-and-how-i-built-a-better-one-with-claude.jsonld"}}