pgvector

Curso System Design Interview Notebook LM - From Zero to Production

Open-source, local-first alternative to NotebookLM. Chat with PDFs, links, YouTube videos and Arabic articles; every answer cites the exact page or moment, and flashcards, quizzes and mind maps open their source. Runs offline on Ollama, or with OpenRouter, NVIDIA, OpenAI, Anthropic, Google or Cohere.

Upload sources, chat with them, and trace every statement in an answer back to the passage it came from. Next.js, NestJS, Supabase/pgvector, hybrid search with RRF.

A self-contained Google Colab Notebook that implements RAG using Postgres pgvector, pgai, and LLMs. Supports local-Ollama/OpenAI/Anthropic

Local-first AI notebook and knowledge base built with FastAPI, PostgreSQL/pgvector and Vue 3, featuring semantic search, RAG chat, AI writing, quizzes and mind maps. | RAGNotebook-Plus 是一个本地优先的智能笔记与知识库系统，支持资料上传解析、语义检索、RAG 问答、AI 写作辅助、快速测试与思维导图。

Open-source AI reading studio for grounded document chat, semantic search, notebooks, audio, and social discovery.

Eval-first notebooks for CPT, SFT, DPO and RAG on a 0.5B model, on a laptop. Fictional data included, outputs stripped.

NotebookLM-style RAG app: notebook-isolated pgvector search over PDFs, web pages, YouTube and transcripts, with streamed answers whose citations jump to the exact page, timestamp or passage.

One notebook that every AI agent you use can read, on every machine you own. Shared markdown memory for Codex, Claude Code and Gemini, with optional MCP search.

End-to-end customer-support LLM: LoRA fine-tune Llama 3.2 3B on Bitext, quantize to Q4 GGUF (Ollama), RAG with pgvector, serve via FastAPI. Includes benchmarks, ablations, and Kaggle notebook.

Capture-first research notebook (a modern commonplace book / second brain). Drop a topic in one line; Neemee researches it in the background and files it back as a structured, connected brief that gets smarter as it grows.

Free AI/ML interview practice built on your own study material. Daedalus writes conceptual interview questions from your PDFs, Jupyter notebooks and arXiv papers, grades your answers against those same sources with cited feedback, and schedules reviews with FSRS spaced repetition. FastAPI, Postgres + pgvector, local and free-tier LLMs.