Recently Added
3,528 documents
DevFlow Cognitive Task+Memory System - Project Context
Documents a TypeScript cognitive framework with hierarchical task management, SQLite-based persistent memory, and multi-platform AI orchestration.
Building a RAG Pipeline from Scratch
Teaches how to build a modular RAG pipeline from scratch with chunking, embeddings, vector storage, retrieval, and generation.
Development Summary
Documents the development journey of an AI agent server, covering AI-assisted vs hand-written code, bugs, architecture, and production plans.
SMART-CAM (Smart AI Camera) — Full Documentation
Documents an edge AI surveillance pipeline that detects people, extracts ReID embeddings, and tracks identities locally on a Raspberry Pi with optional Hailo NPU acceleration.
Enso Atlas: Technical Specification
Describes a modular, on-premise pathology AI platform with swappable foundation models, classification heads, and config-driven cancer-type projects.
Pulse — Life Cofounder | Build Log
Documents a full-stack monorepo that ingests LinkedIn and GitHub data, generates embeddings in-browser, and provides a RAG chat with an AI cofounder.
ReachInbox Email Aggregator - Development Plan
Breaks a full-stack email aggregator into 6 sequential phases with mandatory, core, and stretch tiers for an interview project.
Self-Improving Job Applier — Technical Implementation Spec
Defines a self-improving job application pipeline that scrapes Seek, classifies job descriptions into four archetypes, tracks outcomes via Gmail, and triggers resume rewrites based on market drift.
AI Research Assistant Usage Guide
Guides researchers through installing and using a tool that combines knowledge graphs, semantic search, and AI agents for paper analysis.
Document DB (MCP Server) Requirements Specification
Specifies a local vector server for Japanese documents using OpenAI embeddings and sqlite-vec, exposed as MCP tools for insert/find/delete.
Startup Ecosystem Platform — Implementation Plan
Breaks a full-stack startup platform into 10 ordered phases with checklists for backend, AI, and frontend implementation.
scope
Defines a Python tool that downloads YouTube transcripts and comments, then runs a 3-stage AI pipeline to produce topic summaries and atomic insights with hybrid search.
Vesper — Implementation Plan
Breaks a journal app build into 5 sequential phases with checkboxes, exit criteria, and file references for each task.
plan
Maps a full-stack AI learning platform with Next.js 15, PostgreSQL, Gemini API, and RAG across 8 implementation phases.
CricOracle
Plans a RAG system that answers cricket law questions by crawling MCC rules, embedding them, and retrieving relevant chunks before querying an LLM.
AI Scientist Lab Notebook — Architecture (V1)
Defines the full architecture for an AI scientist lab notebook with RAG, ingestion, and plotting pipelines.
Software Requirements Specification (SRS)
Defines functional and non-functional requirements for a RAG-based corporate knowledge assistant with citation and hallucination guardrails.
Contract RAG MVP — Plan
Defines a complete local-only contract analysis MVP with document upload, background processing, and RAG extraction via Ollama.
AI Chatbot Maker with TiDB Vector Search - Hackathon Plan
Outlines a hackathon plan for building an AI chatbot platform that ingests documentation and GitHub repos, uses TiDB vector search, and generates embeddable script tags.
RAG System Documentation
Documents a full RAG pipeline using MongoDB text search and Gemini for AI answers from user-uploaded study materials.
Hallucination Guard - System Documentation & Literature Review
Documents a hallucination detection system with literature review, architecture, and methodology for LLM output verification.
Delivery Shield - Setup Guide
Guides setup of a monorepo refund system with mock and production modes, including MongoDB, Gemini AI, and CDP integration.
Legal Brief Matcher: Technical Documentation
Documents a legal brief matching system that uses embeddings, citation analysis, and multi-factor confidence scoring to align arguments between opposing briefs.
Multilingual Agentic RAG (venv, Qdrant, NiceGUI, Gemini-001)
Defines a multilingual agentic RAG pipeline using Gemini embeddings, Tesseract OCR, Qdrant, and NiceGUI with CPU-first defaults.