Agentic Brain Tutorial Documentation Index

Complete tutorial documentation for building production-ready AI chatbots with persistent memory.

📚 Quick Navigation

Document	Duration	Level	What You'll Learn
Getting Started	5 min	Beginner	Installation, first bot, memory basics
Tutorial 1: Simple Chatbot	15 min	Beginner	Custom personality, multi-turn conversations, error handling
Tutorial 2: Adding Memory	20 min	Intermediate	Neo4j integration, storing facts, semantic search
Tutorial 3: RAG Chatbot	25 min	Intermediate	Document retrieval, embeddings, grounded responses
Tutorial 4: Multi-User SaaS	30 min	Advanced	Tenants, access control, rate limiting, multi-tenancy
Tutorial 5: Production Deployment	20 min	Advanced	Docker, Kubernetes, monitoring, security

Total Time: ~115 minutes (less than 2 hours!)
Total Coverage: From zero to production-ready SaaS chatbot

🎯 Learning Paths

Path 1: "I just want a chatbot" ⚡

Getting Started (5 min)
    ↓
Tutorial 1: Simple Chatbot (15 min)
    ↓
You have a working chatbot!

Path 2: "I want to build a startup" 🚀

Getting Started (5 min)
    ↓
Tutorial 1: Simple Chatbot (15 min)
    ↓
Tutorial 2: Adding Memory (20 min)
    ↓
Tutorial 4: Multi-User SaaS (30 min)
    ↓
Tutorial 5: Production Deployment (20 min)
    ↓
You're ready to launch!

Path 3: "I need intelligent document Q&A" 📚

Getting Started (5 min)
    ↓
Tutorial 3: RAG Chatbot (25 min)
    ↓
You can answer questions about any documents

Path 4: "I want to master everything" 🏆

Follow all tutorials in order (1-5)
    ↓
You're an expert in AI chatbots!

📖 Document Descriptions

Getting Started (5 minutes)

File: getting-started.md

Your first step! This quick start covers:

Installation in 3 commands
Your first chatbot in 30 seconds
How to add persistent memory
Common troubleshooting

Perfect for: Anyone new to Agentic Brain

Prerequisites: Python 3.10+, Docker

Expected outcome: A working chatbot that remembers things between sessions

Tutorial 1: Build Your First Chatbot (15 minutes)

File: tutorials/01-simple-chatbot.md

Create a production-ready chatbot with:

System prompts and personality
Multi-turn conversations
Error handling
Conversation logging
Multiple LLM providers (Ollama, OpenAI, Anthropic)
Multi-user support

Perfect for: Developers building chatbot applications

Prerequisites: Completed Getting Started guide

Expected outcome: A well-structured chatbot you can extend

Code example:

bot = SimpleChatbot(name="assistant", user_id="demo_user_001")
response = bot.chat("Hello! I'm working on a Python project.")
# Bot remembers and references your project in follow-ups

Tutorial 2: Adding Neo4j Memory (20 minutes)

File: tutorials/02-adding-memory.md

Master persistent memory with:

Memory types (facts, skills, preferences, experiences)
Neo4j graph database fundamentals
Semantic search across memories
User profiles from stored data
Data export for GDPR compliance
Memory cleanup and archival

Perfect for: Building knowledge-aware AI systems

Prerequisites: Tutorial 1 completed

Expected outcome: Chatbots that build rich knowledge about users

Code example:

memory = AdvancedMemory(uri="bolt://localhost:7687")
memory.store_memory(UserMemory(
    user_id="alice",
    memory_type=MemoryType.SKILL,
    content="Expert in Python and FastAPI"
))

# Later, retrieve and use
profile = memory.get_user_profile("alice")
print(profile.skills)  # ["Expert in Python and FastAPI"]

Tutorial 3: RAG Chatbot (25 minutes)

File: tutorials/03-rag-chatbot.md

Ground responses in documents with:

Document loading and chunking
Semantic search with embeddings
Retrieval-Augmented Generation (RAG) workflow
Source citation in responses
Knowledge base management
Keyword and semantic search fallback

Perfect for: Building Q&A systems, knowledge base assistants

Prerequisites: Tutorial 1-2 completed

Expected outcome: Chatbot that answers questions accurately based on your documents

Use cases:

Customer support (knowledge base)
Technical documentation
FAQ assistant
Legal document Q&A

Code example:

bot = RAGChatbot(
    name="support_agent",
    user_id="demo_user",
    documents_dir="./knowledge_base",
    use_embeddings=True
)

response = bot.chat("What's your refund policy?")
# Bot retrieves policy from documents and cites source
# Output: "...30-day refunds... (Source: refund_policy.txt)"

Tutorial 4: Multi-User SaaS (30 minutes)

File: tutorials/04-multi-user.md

Build enterprise-grade multi-tenancy with:

Tenant models and data isolation
User roles and access control
Rate limiting per customer
Usage analytics
Billing integration points
FastAPI endpoints with auth
SaaS architecture patterns

Perfect for: Building chatbot platforms, selling AI services

Prerequisites: Tutorials 1-3 completed

Expected outcome: A multi-tenant SaaS chatbot backend

Architecture:

┌─────────────────────────────────────┐
│    Your SaaS API                    │
├─────────────────────────────────────┤
│ Customer A │ Customer B │ Customer C│
│ (isolated) │ (isolated) │ (isolated)│
└─────────────────────────────────────┘
     │           │           │
     └───────────┴───────────┘
            ↓
        Shared Database
        (Neo4j + Redis)

Code example:

# Create tenants
customer_a = tenant_manager.create_tenant("acme_corp", "ACME", tier=TenantTier.PRO)
customer_b = tenant_manager.create_tenant("widgets_inc", "Widgets", tier=TenantTier.FREE)

# Chat with isolation
response = chatbot.chat(
    tenant_id="acme_corp",
    user_id="alice",
    message="Hello!"
)  # Alice's data is isolated from Widgets' users

Tutorial 5: Production Deployment (20 minutes)

File: tutorials/05-deployment.md

Deploy to production with:

Multi-stage Docker builds
Docker Compose full stack
Kubernetes manifests
Health checks and readiness probes
Prometheus + Grafana monitoring
Security hardening
Zero-downtime deployments
Production backup strategies

Perfect for: DevOps, platform engineers, production readiness

Prerequisites: Tutorials 1-4 completed

Expected outcome: Production-ready deployment with monitoring

Stack includes:

Neo4j (graph database)
Redis (caching)
Ollama (local LLM)
Prometheus (metrics)
Grafana (dashboards)

Quick start:

docker-compose up -d
# All services online with health checks

curl http://localhost:8000/health
# {"status": "ok", "version": "0.1.0"}

🔑 Key Concepts Covered

Core Architecture

Neo4j: Graph database for persistent knowledge
Sessions: User conversation state
Memory: Facts, preferences, experiences
Embeddings: Semantic search via sentence transformers
RAG: Retrieval-Augmented Generation for grounding

LLM Providers

Ollama (local, open-source)
OpenAI (GPT-3.5, GPT-4)
Anthropic (Claude)
Custom providers via LLMRouter

Deployment

Docker containerization
Docker Compose orchestration
Kubernetes scaling
Health checks and monitoring
Production security

Enterprise Features

Multi-tenancy with data isolation
Role-based access control (RBAC)
Rate limiting and quotas
Usage analytics
GDPR data export
Audit logging

📋 Prerequisites Progression

Tutorial	Requires	Time to Setup
Getting Started	Python 3.10+, Docker	~5 min
Tutorial 1	Getting Started complete	0 min
Tutorial 2	Tutorial 1 complete	0 min
Tutorial 3	sentence-transformers	~5 min (first run slow)
Tutorial 4	FastAPI (pip install)	~2 min
Tutorial 5	Docker Compose	~1 min

🛠️ Common Use Cases & Which Tutorial

Use Case	Start Here	Why
Chatbot for my website	Tutorial 1	Simple, self-contained
Support bot that learns	Tutorial 2	Memory integration key
Q&A over my docs	Tutorial 3	RAG perfect for this
White-label chatbot SaaS	Tutorial 4	Multi-tenancy required
Production deployment	Tutorial 5	Scale and monitor
All of the above	Tutorials 1-5	Full stack

🆘 Troubleshooting

By Tutorial

Getting Started Issues? → See "Troubleshooting" section in getting-started.md

Tutorial 1 Issues? → See "Troubleshooting" section in 01-simple-chatbot.md

Tutorial 2 Issues? → See "Troubleshooting" section in 02-adding-memory.md

Tutorial 3 Issues? → See "Troubleshooting" section in 03-rag-chatbot.md

Tutorial 4 Issues? → See "Troubleshooting" section in 04-multi-user.md

Tutorial 5 Issues? → See "Troubleshooting" section in 05-deployment.md

Common Issues Across All Tutorials

Neo4j connection refused
→ docker ps | grep neo4j
→ docker run -d -p 7687:7687 -e NEO4J_AUTH=neo4j/password neo4j:latest

Ollama not responding
→ ollama serve (if local) or use OpenAI/Anthropic

Python version too old
→ python3 --version should be 3.10+

Import errors
→ pip install -r requirements.txt

💡 Pro Tips

1. Development vs Production

Dev: Use Ollama locally (free, no API keys)
Prod: Use cloud LLMs (faster, better models) or large Ollama instances

2. Memory Best Practices

Always pass user_id for proper isolation
Don't store sensitive data in memories (compliance!)
Regularly archive old conversations
Use confidence field to weight memories

3. RAG Best Practices

Test with your actual documents first
Chunk size of 500-1000 characters usually works well
Enable embeddings for better retrieval
Monitor query performance

4. Multi-Tenancy Best Practices

Always scope user IDs by tenant: f"{tenant_id}#{user_id}"
Enforce rate limits strictly
Audit all administrative actions
Test data isolation thoroughly

5. Production Best Practices

Use Docker Compose locally, Kubernetes in prod
Monitor CPU/memory constantly
Set up alerting for service failures
Test disasters recovery regularly

📚 Additional Resources

Official Documentation

Related Tools & Libraries

sentence-transformers - Embeddings
ChromaDB - Vector database
LangChain - LLM orchestration
FastAPI - API framework

Learning Resources

🎓 Your Learning Journey

START HERE
    ↓
[Getting Started] - 5 minutes
    ↓
[Tutorial 1] Simple Chatbot - 15 minutes
    ↓
[Tutorial 2] Memory - 20 minutes
    ✓ You can build chatbots with memory!
    ↓
[Tutorial 3] RAG - 25 minutes
    ✓ You can answer Q&A over documents!
    ↓
[Tutorial 4] SaaS - 30 minutes
    ✓ You can serve multiple customers!
    ↓
[Tutorial 5] Deployment - 20 minutes
    ✓ You can run in production!
    ↓
YOU ARE AN AGENTIC BRAIN EXPERT! 🎉

🤝 Contributing

Found an issue or improvement?

Check existing docs for answers
File issues on GitHub
Submit PRs for improvements
Share your use cases in discussions

📞 Support

Questions about a specific tutorial?

Check the troubleshooting section in that tutorial
Look for the PRE-REQUISITES section to verify setup

General questions?

Check the main README
See the FAQ if it exists
Open an issue on GitHub

Having trouble following along?

Start with Getting Started, don't skip steps
Make sure each prerequisite is installed
Run health checks frequently
Check service logs: docker-compose logs service_name

📄 License

All tutorials and documentation are part of Agentic Brain, licensed under Apache-2.0.

Ready to build? Start with Getting Started →

Want to contribute? See CONTRIBUTING.md

Have feedback? Open an issue

Made with ❤️ for developers building intelligent systems

Getting Started • Tutorials • README • Issues

FilesExpand file tree

README.md

Latest commit

History

README.md

File metadata and controls

Agentic Brain Tutorial Documentation Index

📚 Quick Navigation

🎯 Learning Paths

Path 1: "I just want a chatbot" ⚡

Path 2: "I want to build a startup" 🚀

Path 3: "I need intelligent document Q&A" 📚

Path 4: "I want to master everything" 🏆

📖 Document Descriptions

Getting Started (5 minutes)

Tutorial 1: Build Your First Chatbot (15 minutes)

Tutorial 2: Adding Neo4j Memory (20 minutes)

Tutorial 3: RAG Chatbot (25 minutes)

Tutorial 4: Multi-User SaaS (30 minutes)

Tutorial 5: Production Deployment (20 minutes)

🔑 Key Concepts Covered

Core Architecture

LLM Providers

Deployment

Enterprise Features

📋 Prerequisites Progression

🛠️ Common Use Cases & Which Tutorial

🆘 Troubleshooting

By Tutorial

Common Issues Across All Tutorials

💡 Pro Tips

1. Development vs Production

2. Memory Best Practices

3. RAG Best Practices

4. Multi-Tenancy Best Practices

5. Production Best Practices

📚 Additional Resources

Official Documentation

Related Tools & Libraries

Learning Resources

🎓 Your Learning Journey

🤝 Contributing

📞 Support

📄 License