GuidedMindGuidedMind
Hybrid Search · GraphRAG · MCP · Python SDK · Eval Benchmarking · Monitoring

Giving your AI context is the easy part. Improving it isn't.

GuidedMind adds the iteration layer your pipeline is missing — eval benchmarking, live query monitoring, and hybrid search you can tune without touching code.

Free tier — no credit card required

SOC 2 Type II — In Progress
LS
Lumora Support Bot
Your client's agent
MCP
Can I return the X200 after 20 days?
Yes, you can return it within 30 days of purchase.
✕ Wrong. Bundle items: 14 days.
New project · Lumora support
returns_policy.md
Markdown
product_manual.pdf
Scanned, 2 columns
pricing_2025.xlsx
Spreadsheet
onboarding.docx
Word
AI assistant
Set this up as a support bot for Lumora's customers.
Chunking:document-based
Search:hybrid
Knowledge graph extraction:normal
Graph extraction level: set here.
Metadata:product, version
4.2 Batt ery & warr 5.1 Clean ing The X200 ships wi th a Use a dry cloth on ly. Do rechargeable Li-ion not immerse the Model Capacity Runtime X200 2400mAh 7h X200 Pro 3200 mAh 9h battery. Charge fully be fore first use. The warr anty cov ers 2 yrs fr om purch ase ||| 12
Raw extraction · product_manual.pdf p.12
Parsed · product_manual.pdf p.12
4.2 Battery & warranty
The X200 ships with a rechargeable Li-ion battery. Charge fully before first use. The warranty covers 2 years from purchase.
Model
Capacity
Runtime
X200
2400 mAh
7 h
X200 Pro
3200 mAh
9 h
Bad chunk cuts cause wrong answers.
5.1 Cleaning
Use a dry cloth only. Do not immerse the device.

GraphRAG Knowledge Graphs

Automatically extract entities and relationships from documents. Cross-document linking, smart graph editor, and configurable extraction depths.

Hybrid Search

Combine dense vector search with sparse BM25 retrieval. Metadata filtering, similarity thresholds, and cross-encoder re-ranking for optimal results.

RAG Configurator

Full control over chunking strategies, embedding models, and retrieval pipelines. Clone any project in one click — fork a working config as your starting point.

API, SDK & MCP

Connect via REST API, Python SDK, or MCP server. Works natively with Claude, Cursor, and Windsurf.

pip install guidedmind-rag-sdk-python

Eval & Benchmarking

Compare two projects side-by-side. Measure retrieval quality, latency, and ranking accuracy. Know your config improved — no guessing.

Architecture: The RAG Pipeline

Click on any stage below to explore our advanced processing capabilities, from ingestion to generation.

Intelligent Document Processing

Our ingestion engine handles unstructured data from diverse sources including direct File Uploads, URLs, and Google Drive integrations. We go beyond simple text extraction.

Key Features

PDF, MD, TXT, CSV, XLS, DOCX Support
Tesseract OCR Engine for scanned docs
ML-based Layout Analysis (Column/Paragraph detection)
Special symbol cleaning & sanitization

Flexible Architecture

One platform, two powerful deployment modes. Choose the integration that fits your stack.

Semantic Search Engine

Retrieval Only

Use our high-performance vector API to power search bars, recommendation feeds, or existing applications. We return ranked document chunks with relevance scores.

  • Low-latency retrieval
  • Hybrid keyword + vector ranking
  • Raw metadata access
GET /api/v1/search?q=legal_precedent&top_k=5

AI Agent (Context Engine)

Retrieval + Generation

Deploy a full reasoning engine. The system retrieves context, ranks it, and uses an LLM to synthesize a natural language answer with citations.

  • Multi-turn conversation history
  • Grounded answers with citations
  • Short & Long Memory built in
POST /api/v1/agent/chat { "message": "..." }

Python SDK

Type-safe Client

Type-safe client with async support, Short & Long Memory, auto-retry. Get started in minutes with full IDE support.

pip install guidedmind-rag-sdk-python
View on GitHub

n8n Integration

Workflow Automation

Use GuidedMind RAG inside your n8n workflows. Official node available — drag-and-drop document ingestion and hybrid search for any automation.

View n8n node

Built for Business

From legal firms to e-commerce giants, our platform adapts to your industry's data.

Legal Knowledge Base

Law firms and legal departments

Index your contracts, case law and legal texts. Get precise answers with source citations. Save hours of document research.

E-commerce Chatbot

Online stores and marketplaces

Create an assistant that knows your entire product catalog. Personalized recommendations, technical Q&A, 24/7 sales support.

HR & Onboarding

Human resources departments

Centralize your internal procedures, policies, and handbooks. New employees autonomous from day one. Instant answers to common questions.

Customer Support

After-sales and helpdesk services

Technical knowledge base accessible in natural language. Reduced resolution time. Better-equipped agents with the right information.

Research & Monitoring

R&D teams and analysts

Explore thousands of research documents, patents, articles. Identify trends and connect information across different sources.

AI Agent Builder

Developers building stateful agents

Build stateful agents with built-in Short Memory (session) and Long Memory (user context). Your agent remembers what was said — and recalls it across sessions.

Desktop App

Run it on your own machine

Point the desktop client at a local model and keep every document, conversation and generated file on the machine it was created on. Each chat gets its own sandbox, and nothing leaves it.

Download for macOSDownload for Windows

macOS 13+ · Windows 10+ · Linux coming soon

Production-Grade Security

Your data security is our top priority. We're committed to the highest standards of compliance and protection.

End-to-End Encryption

All data encrypted in transit and at rest using industry-standard AES-256 encryption.

SOC-2 Compliance Badge

SOC 2 Compliance

🔒 SOC 2 Type II — In Progress

Expected Q2/Q3 2026

Secure Infrastructure

Hosted on enterprise cloud infrastructure with continuous monitoring and automatic failover.