{"slug": "10th-k-ai-i-built-an-ai-tutor-for-my-sister-who-was-stuck-on-textbook-questions", "title": "10th K AI: I Built an AI Tutor for My Sister Who Was Stuck on Textbook Questions", "summary": "A developer built 10th K AI, a retrieval-augmented generation tutor for Karnataka SSLC Class 10 textbooks, created for a sister who struggled to parse textbook question phrasing. The system chunks textbook PDFs, embeds them with Sentence Transformers' all-MiniLM-L6-v2 into a FAISS index, and combines keyword and semantic hybrid retrieval with an open-weight LLM served locally via Ollama to produce answers grounded in textbook evidence, along with question-pattern recognition and memory aids.", "body_md": "This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend\n\n10th K AI — An AI Tutor I Built for My Sister\n\nMy sister is a Class 10 student.\n\nI noticed something frustrating when she was studying: sometimes she knew the subject, but still got stuck because the textbook language was difficult to understand.\n\nThe bigger problem wasn't always \"I don't know the answer.\"\n\nIt was:\n\n\"I don't understand what type of question this is asking me.\"\n\nA question might be asking for a person and their contribution, a cause and its effect, a formula and a calculation, or a definition and a term.\n\nSo I built 10th K AI specifically for her.\n\nIt is an AI-powered Karnataka SSLC Class 10 textbook tutor that doesn't just give an answer. It helps a student understand:\n\nWhat is the concept? → What question pattern is this? → How do I solve it? → How do I remember it?\n\nWhat I Built\n\n10th K AI lets a student ask a question from their Class 10 textbook and get a response grounded in the actual textbook.\n\nFor example, if the student asks:\n\nWho derived the quadratic formula?\n\n10th K AI searches the Karnataka SSLC textbook and finds the relevant passage:\n\n\"Sridharacharya (C.E. 1025) derived a formula, now known as the quadratic formula...\"\n\nInstead of stopping there, the tutor turns the information into a learning structure:\n\nAnswer\n\n Textbook Evidence\n\n Simple Explanation\n\n Question Pattern\n\n Memory Trick\n\n Similar Practice Question\n\nThe goal is not just to help a student get one answer.\n\nThe goal is to help them recognize the pattern the next time they see a similar question.\n\nWho I Built It For\n\nI built 10th K AI for my sister.\n\nWhile helping her study, I realized that students can struggle even when the information is technically available to them.\n\nA textbook may say something like:\n\n\"Sridharacharya derived a formula...\"\n\nBut an exam question might ask:\n\n\"Who derived the quadratic formula?\"\n\nThe student needs to recognize:\n\nPerson → Contribution\n\nThat small recognition can make a big difference during an exam.\n\nThat became the central idea behind 10th K AI.\n\nHow It Works\n\n10th K AI uses a Retrieval-Augmented Generation (RAG) pipeline.\n\nThe textbook PDFs are processed into smaller chunks.\n\nEach chunk stores information such as:\n\nSubject\n\nBook\n\nPage number\n\nTextbook content\n\nI created embeddings for the textbook chunks using:\n\nSentence Transformers — all-MiniLM-L6-v2\n\nThose embeddings are stored in a FAISS vector index.\n\nWhen a student asks a question:\n\nStudent Question\n\n       ↓\n\nKeyword Search\n\n       ↓\n\nSemantic Search\n\n       ↓\n\nHybrid Retrieval\n\n       ↓\n\nRelevant Textbook Passage\n\n       ↓\n\nOpen-Source LLM\n\n       ↓\n\nLearning-focused Answer\n\nThe LLM is instructed to use the retrieved textbook evidence rather than simply inventing an answer from general knowledge.\n\nOpen-Source AI at the Core\n\nThe project uses Ollama and an open-weight model for the AI generation layer.\n\nThis was important to me because I wanted the project to be built around an AI system that I could actually control and run locally.\n\nWith Ollama, I can run the model on my own machine during development instead of making my application completely dependent on a closed AI API.\n\nThe architecture also keeps the LLM layer replaceable, which means I can experiment with different open models without rebuilding the entire RAG system.\n\nWhy Open Innovation Matters\n\nFor an educational tool, I think control over the AI layer matters.\n\nWith a closed API, the model is essentially a service I depend on.\n\nWith an open model through Ollama, I can:\n\nRun inference locally\n\nExperiment with different models\n\nControl the prompting\n\nChange the retrieval strategy\n\nKeep the textbook retrieval pipeline under my control\n\nBuild the system without locking the entire project to one proprietary model\n\nFor 10th K AI, the combination of open model + local inference + textbook RAG makes the project much more understandable and controllable.\n\nThe Interesting Part: Question Patterns\n\nThis is the part I care about most.\n\nI don't want 10th K AI to become another chatbot where a student asks:\n\n\"What is the answer?\"\n\nand immediately copies it.\n\nInstead, I want the student to gradually recognize patterns.\n\nFor example:\n\nCause → Effect\n\nEvent → Person\n\nDefinition → Term\n\nFormula → Numerical calculation\n\nThe student starts learning:\n\n\"I've seen this type of question before.\"\n\nThat is the skill I wanted to build into the tutor.\n\nTech Stack\n\nFrontend / App\n\nStreamlit\n\nAI\n\nOllama\n\nOpen-weight LLM\n\nRAG\n\nFAISS\n\nSentence Transformers\n\nall-MiniLM-L6-v2\n\nText Processing\n\nPyMuPDF\n\nPython\n\nData\n\nKarnataka SSLC Class 10 English-medium textbooks\n\nDeployment\n\nStreamlit Community Cloud\n\n Demo\n\nLive Demo:\n\n[https://10thk-ai-d6w3ycqcfe6i4zdecucqxp.streamlit.app/](https://10thk-ai-d6w3ycqcfe6i4zdecucqxp.streamlit.app/)\n\nTry asking:\n\nor another question from the Karnataka SSLC Class 10 Mathematics, Science, or Social Science textbooks.\n\nCode\n\nGitHub:\n\n[https://github.com/aditikshetty/10thK-AI](https://github.com/aditikshetty/10thK-AI)\n\nThe repository contains the RAG pipeline, application code, textbook processing code, and retrieval system.\n\nWhy I Built It\n\nI didn't start this project because I wanted to build another AI chatbot.\n\nI started because I saw someone I care about getting stuck while studying.\n\nThat made me think about what an AI tutor should actually do.\n\nMaybe the most useful answer isn't always:\n\n\"Here is the answer.\"\n\nSometimes it is:\n\n\"This is the kind of question you're looking at. Here's how you recognize it next time.\"\n\nThat's what I wanted 10th K AI to teach.\n\nWhat's Next\n\nThere is still a lot I want to improve:\n\nBetter question-pattern classification\n\nSupport for textbook diagrams and images\n\nMore accurate similar-question generation\n\nStudent progress tracking\n\nPersonalized practice sets\n\nVoice-based interaction\n\nMore regional-language support\n\nBetter evaluation of RAG accuracy\n\nBut the first version is now working, and more importantly, it was built for a real person with a real problem.\n\nWhat I Learned\n\nThe biggest lesson from this weekend wasn't just about RAG or embeddings.\n\nIt was that a small, specific problem can be much more meaningful than a generic AI idea.\n\nInstead of asking:\n\n\"What AI app can I build?\"\n\nI started with:\n\n\"What is frustrating someone I care about, and can I build something to help?\"\n\nThat changed the entire project.\n\nPrize Categories\n\nI am submitting 10th K AI for the overall Hacktoberfest Weekend Challenge: Build for a Friend.", "url": "https://wpnews.pro/news/10th-k-ai-i-built-an-ai-tutor-for-my-sister-who-was-stuck-on-textbook-questions", "canonical_source": "https://dev.to/aditi_shetty_caaab207ff98/10th-k-ai-i-built-an-ai-tutor-for-my-sister-who-was-stuck-on-textbook-questions-40eb", "published_at": "2026-10-04 22:02:45+00:00", "updated_at": "2026-10-04 22:12:17.536927+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "generative-ai", "ai-tools", "natural-language-processing"], "entities": ["10th K AI", "Ollama", "FAISS", "Sentence Transformers", "all-MiniLM-L6-v2", "Karnataka SSLC"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/10th-k-ai-i-built-an-ai-tutor-for-my-sister-who-was-stuck-on-textbook-questions", "markdown": "https://wpnews.pro/news/10th-k-ai-i-built-an-ai-tutor-for-my-sister-who-was-stuck-on-textbook-questions.md", "text": "https://wpnews.pro/news/10th-k-ai-i-built-an-ai-tutor-for-my-sister-who-was-stuck-on-textbook-questions.txt", "jsonld": "https://wpnews.pro/news/10th-k-ai-i-built-an-ai-tutor-for-my-sister-who-was-stuck-on-textbook-questions.jsonld"}}