Skip to content

Commit 0d7669b

Browse files
NabidAlamNabidAlam
committed
Restructure 25-generative-ai-llms module to match other chapters
- Update README.md to follow standard module structure: - Phase 7 title and proper formatting - 'What You'll Learn' section - 'Topics Covered' with detailed breakdown (6 main topics) - 'Learning Objectives' section - 'Projects' section with 5 project ideas - 'Key Concepts' section - 'Documentation & Learning Resources' with official docs, courses, tutorials, videos, practice - Navigation links (Previous Phase, Next Phase) - Structure now matches modules like 12-natural-language-processing and 11-computer-vision - Maintains links to comprehensive guides in resources/ folder
1 parent c9d43e0 commit 0d7669b

File tree

1 file changed

+154
-166
lines changed

1 file changed

+154
-166
lines changed

‎25-generative-ai-llms/README.md‎

Lines changed: 154 additions & 166 deletions
Original file line numberDiff line numberDiff line change
@@ -1,169 +1,157 @@
1-
# Generative AI & Modern LLM Applications
2-
3-
Complete guide to modern Large Language Model (LLM) applications, prompt engineering, vector databases, RAG systems, and building production-ready GenAI applications.
4-
5-
## Overview
6-
7-
This module covers the modern approach to NLP and AI applications using pre-trained foundational models (LLMs) rather than training models from scratch. This is the current industry standard for most NLP and AI applications.
8-
9-
## What You'll Learn
10-
11-
- **Prompt Engineering**: Comprehensive techniques for effective LLM interaction
12-
- **Vector Databases**: Semantic search and similarity matching with Pinecone, ChromaDB, Weaviate, FAISS
13-
- **RAG (Retrieval-Augmented Generation)**: End-to-end implementation for knowledge-augmented LLMs
14-
- **LLM Agents**: Building autonomous AI agents with LangChain, LangGraph, AutoGPT
15-
- **Multi-Agent Systems**: Coordinated workflows with multiple specialized agents
16-
- **Building Production GenAI Apps**: Real-world deployment patterns and best practices
17-
18-
## Prerequisites
19-
20-
Before starting this module, you should have completed:
21-
- **Phase 6: Specialized Deep Learning** (Modules 11-12, 15)
22-
- Understanding of Transformers (Module 12)
23-
- Basic NLP concepts (Module 12)
24-
- Python programming (Module 00-01)
25-
26-
**Note**: This module can also be learned in parallel with Module 12 (NLP) if you want to learn modern approaches early.
27-
28-
## Module Structure
29-
30-
### Core Topics
31-
32-
1. **Prompt Engineering** (`prompt-engineering.md`)
33-
- What is prompt engineering and why it matters
34-
- Zero-shot vs few-shot prompting
35-
- Chain-of-thought reasoning
36-
- Handling AI hallucinations
37-
- Text embeddings and vectors
38-
- Advanced prompting techniques
39-
- Prompt engineering for different tasks
40-
41-
2. **Vector Databases** (`vector-databases.md`)
42-
- What are vector databases and why they're essential
43-
- Pinecone: Managed vector database
44-
- ChromaDB: Open-source vector database
45-
- Weaviate: Vector search engine
46-
- FAISS: Facebook AI Similarity Search
47-
- Semantic search and similarity matching
48-
- Choosing the right vector database
49-
50-
3. **RAG (Retrieval-Augmented Generation)** (`rag-systems.md`)
51-
- RAG architecture and components
52-
- Document ingestion and chunking
53-
- Embedding generation
54-
- Vector database integration
55-
- Query processing and retrieval
56-
- LLM integration for generation
57-
- Evaluation metrics for RAG systems
58-
- Production deployment patterns
59-
60-
4. **LLM Agents** (`llm-agents.md`)
61-
- What are AI agents
62-
- LangChain for building agents
63-
- LangGraph for complex agent workflows
64-
- AutoGPT and autonomous agents
65-
- Tools and function calling
66-
- Memory and context management
67-
- Agent evaluation
68-
69-
5. **Multi-Agent Systems** (`multi-agent-systems.md`)
70-
- Multi-agent architectures
71-
- Agent coordination and communication
72-
- Specialized agent roles (Planner, Research, Writer)
73-
- CrewAI and AutoGen frameworks
74-
- MCP (Model Context Protocol)
75-
- A2A (Agent-to-Agent) communication
76-
77-
6. **Building Production GenAI Apps** (`production-genai.md`)
78-
- Tech stack for GenAI applications
79-
- Streamlit for GenAI interfaces
80-
- FastAPI for GenAI backends
81-
- Deployment strategies
82-
- Cost optimization
83-
- Monitoring and observability
84-
- Security best practices
85-
86-
## Comprehensive Guides
87-
88-
For detailed coverage of all topics, see:
89-
90-
- **[Generative AI Comprehensive Guide](../resources/generative_ai_comprehensive_guide.md)** - Complete overview with all concepts
91-
- **[RAG Comprehensive Guide](../resources/rag_comprehensive_guide.md)** - Deep dive into RAG implementation
92-
- **[LangChain Guide](../resources/langchain_guide.md)** - LangChain framework details
93-
- **[AI Agents Guide](../resources/ai_agents_guide.md)** - AI agents and multi-agent systems
94-
95-
## Learning Path
96-
97-
### Week 1-2: Prompt Engineering & Vector Databases
98-
- Master prompt engineering techniques
99-
- Set up and use vector databases
100-
- Build semantic search systems
101-
102-
### Week 3-4: RAG Systems
103-
- Implement end-to-end RAG pipeline
104-
- Work with document processing
105-
- Evaluate RAG performance
106-
107-
### Week 5-6: LLM Agents
108-
- Build autonomous agents with LangChain
109-
- Create multi-agent systems
110-
- Deploy agent-based applications
111-
112-
### Week 7-8: Production Deployment
113-
- Deploy GenAI apps to production
114-
- Optimize costs and performance
115-
- Monitor and maintain systems
116-
117-
## Projects
118-
119-
### Project 1: RAG System for Document QA
120-
Build a system that ingests PDFs, stores embeddings in Pinecone/ChromaDB, and answers questions using GPT-4/Llama 3.
121-
122-
**Skills**: RAG, Vector Databases, LLM Integration, LangChain
123-
124-
### Project 2: LLM-Powered Research Agent
125-
Create an autonomous agent that researches topics, gathers information, and generates reports.
126-
127-
**Skills**: LLM Agents, LangGraph, Tool Integration, Multi-step Reasoning
128-
129-
### Project 3: Multi-Agent System
130-
Build a system with specialized agents (Planner, Research, Writer) working together.
131-
132-
**Skills**: Multi-Agent Systems, Agent Coordination, CrewAI/AutoGen
133-
134-
## Time Estimate
135-
136-
- **Full-Time (30-40 hrs/week)**: 1-2 months
137-
- **Part-Time (10-15 hrs/week)**: 2-4 months
138-
139-
## Career Relevance
140-
141-
This module is essential for:
142-
- **LLM Engineer**: Core skills for LLM application development
143-
- **GenAI Solution Architect**: Building production GenAI systems
144-
- **AI Engineer**: Modern AI application development
145-
- **ML Engineer**: GenAI deployment and optimization
146-
147-
## Resources
148-
149-
### Official Documentation
150-
- [LangChain Documentation](https://python.langchain.com/)
151-
- [LangGraph Documentation](https://langchain-ai.github.io/langgraph/)
152-
- [Pinecone Documentation](https://docs.pinecone.io/)
153-
- [ChromaDB Documentation](https://docs.trychroma.com/)
154-
- [Hugging Face Transformers](https://huggingface.co/docs/transformers/)
155-
156-
### Learning Resources
157-
- See [Generative AI Comprehensive Guide](../resources/generative_ai_comprehensive_guide.md) for complete resource list
158-
159-
## Next Steps
160-
161-
After completing this module:
162-
1. Build a RAG system project (see Advanced Projects)
163-
2. Move to Phase 8: Production & MLOps
164-
3. Deploy your GenAI application
165-
4. Continue with specialized topics (Phase 10)
1+
# Phase 7: Generative AI & Modern LLM Applications
2+
3+
Learn to build modern AI applications using Large Language Models (LLMs), prompt engineering, vector databases, RAG systems, and AI agents.
4+
5+
## What You'll Learn
6+
7+
- Prompt Engineering (Zero-shot, Few-shot, Chain-of-Thought)
8+
- Vector Databases (Pinecone, ChromaDB, Weaviate, FAISS)
9+
- RAG (Retrieval-Augmented Generation) Systems
10+
- LLM Agents (LangChain, LangGraph, AutoGPT)
11+
- Multi-Agent Systems (CrewAI, AutoGen)
12+
- Building Production GenAI Apps
13+
- Generative Configuration Parameters
14+
- Model Evaluation and Benchmarks
15+
16+
## Topics Covered
17+
18+
### 1. Prompt Engineering
19+
- **What is Prompt Engineering**: Designing effective inputs for LLMs
20+
- **Zero-shot Prompting**: No examples, rely on pre-trained knowledge
21+
- **Few-shot Prompting**: Provide examples to guide behavior
22+
- **Chain-of-Thought**: Step-by-step reasoning
23+
- **Generative Configuration**: Temperature, top-p, top-k, repetition penalty
24+
- **Handling Hallucinations**: Mitigation strategies
25+
- **Text Embeddings**: Vector representations for semantic search
26+
- **Advanced Techniques**: Role-playing, output formatting, constraints
27+
28+
### 2. Vector Databases
29+
- **What are Vector Databases**: Storage for high-dimensional embeddings
30+
- **Pinecone**: Managed cloud vector database
31+
- **ChromaDB**: Open-source, Python-first vector database
32+
- **Weaviate**: GraphQL-based vector search engine
33+
- **FAISS**: Facebook AI Similarity Search library
34+
- **Semantic Search**: Finding similar documents by meaning
35+
- **Similarity Metrics**: Cosine similarity, Euclidean distance
36+
- **Choosing the Right Database**: Comparison and use cases
37+
38+
### 3. RAG (Retrieval-Augmented Generation)
39+
- **RAG Architecture**: Retrieval + Augmentation + Generation
40+
- **Document Ingestion**: Loading and processing documents
41+
- **Text Chunking**: Strategies for splitting documents
42+
- **Embedding Generation**: Creating vector representations
43+
- **Vector Database Integration**: Storing and retrieving embeddings
44+
- **Query Processing**: User query to embedding conversion
45+
- **Context Augmentation**: Combining retrieved context with prompts
46+
- **LLM Integration**: Generating responses with augmented context
47+
- **Evaluation Metrics**: RAG-specific evaluation methods
48+
- **Production Patterns**: Deployment and optimization strategies
49+
50+
### 4. LLM Agents
51+
- **What are AI Agents**: Autonomous systems that perceive, reason, and act
52+
- **LangChain Agents**: Building agents with LangChain
53+
- **LangGraph**: Graph-based agent workflows
54+
- **AutoGPT**: Fully autonomous goal completion
55+
- **Tools and Function Calling**: Integrating external tools
56+
- **Memory and Context**: Managing conversation history
57+
- **ReAct Framework**: Reasoning and Acting for tool use
58+
- **PAL (Program-aided Language Models)**: Code generation for precise problem solving
59+
- **Agent Evaluation**: Measuring agent performance
60+
61+
### 5. Multi-Agent Systems
62+
- **Multi-Agent Architectures**: Coordinated agent workflows
63+
- **Agent Coordination**: Communication and task distribution
64+
- **Specialized Roles**: Planner, Research, Writer agents
65+
- **CrewAI**: Framework for role-playing agents
66+
- **AutoGen**: Conversational multi-agent systems
67+
- **MCP (Model Context Protocol)**: Standardized context sharing
68+
- **A2A Communication**: Agent-to-agent protocols
69+
70+
### 6. Building Production GenAI Apps
71+
- **Tech Stack**: Frontend, backend, LLM frameworks, vector databases
72+
- **Streamlit**: Fast Python-based UI for GenAI apps
73+
- **FastAPI**: Building GenAI backends
74+
- **Deployment Strategies**: Cloud, on-premise, hybrid
75+
- **Cost Optimization**: Reducing API and infrastructure costs
76+
- **Monitoring and Observability**: Tracking performance and usage
77+
- **Security Best Practices**: API keys, input validation, rate limiting
78+
- **Generative AI Project Lifecycle**: From problem definition to maintenance
79+
80+
## Learning Objectives
81+
82+
By the end of this module, you should be able to:
83+
- Design effective prompts for various LLM tasks
84+
- Set up and use vector databases for semantic search
85+
- Build end-to-end RAG systems for document Q&A
86+
- Create autonomous AI agents with tool integration
87+
- Design and implement multi-agent systems
88+
- Deploy GenAI applications to production
89+
- Optimize LLM applications for cost and performance
90+
- Evaluate and benchmark LLM applications
91+
92+
## Projects
93+
94+
1. **RAG System for Document QA**: Build a system that ingests PDFs, stores embeddings, and answers questions
95+
2. **LLM-Powered Research Agent**: Create an autonomous agent that researches topics and generates reports
96+
3. **Multi-Agent Content Creation**: Build a system with specialized agents (Planner, Research, Writer)
97+
4. **Prompt Engineering Playground**: Experiment with different prompting techniques
98+
5. **Vector Database Comparison**: Compare different vector databases for your use case
99+
100+
## Key Concepts
101+
102+
- **Prompt Engineering**: The art and science of communicating with LLMs
103+
- **Semantic Search**: Finding information by meaning, not keywords
104+
- **RAG**: Combining retrieval with generation for knowledge-augmented AI
105+
- **Agents**: Autonomous systems that can reason and use tools
106+
- **Vector Embeddings**: Dense representations capturing semantic meaning
107+
- **Generative Configuration**: Parameters controlling LLM output (temperature, top-p, etc.)
108+
- **Production Deployment**: Making GenAI apps reliable, scalable, and cost-effective
109+
110+
## Documentation & Learning Resources
111+
112+
**Official Documentation:**
113+
- [LangChain Documentation](https://python.langchain.com/) - Complete LangChain guide
114+
- [LangGraph Documentation](https://langchain-ai.github.io/langgraph/) - Graph-based workflows
115+
- [Pinecone Documentation](https://docs.pinecone.io/) - Managed vector database
116+
- [ChromaDB Documentation](https://docs.trychroma.com/) - Open-source vector database
117+
- [Hugging Face Transformers](https://huggingface.co/docs/transformers/) - Pre-trained models
118+
- [OpenAI API Documentation](https://platform.openai.com/docs) - GPT models and API
119+
120+
**Free Courses:**
121+
- [LangChain Crash Course](https://www.youtube.com/watch?v=lG7Uxts9SXs) - Free YouTube course
122+
- [RAG Tutorial (LangChain)](https://python.langchain.com/docs/use_cases/question_answering/) - Free tutorial
123+
- [Vector Databases Course](https://www.deeplearning.ai/short-courses/vector-databases/) - DeepLearning.AI course
124+
- [Building LLM Applications](https://www.deeplearning.ai/short-courses/building-applications-with-llms/) - DeepLearning.AI course
125+
126+
**Tutorials:**
127+
- [Prompt Engineering Guide](../resources/generative_ai_comprehensive_guide.md#prompt-engineering) - Comprehensive prompt engineering
128+
- [RAG Implementation Guide](../resources/rag_comprehensive_guide.md) - Complete RAG guide
129+
- [LangChain Tutorial](../resources/langchain_guide.md) - LangChain framework guide
130+
- [AI Agents Guide](../resources/ai_agents_guide.md) - Building AI agents
131+
- [Vector Databases Guide](../resources/generative_ai_comprehensive_guide.md#vector-databases) - Vector database comparison
132+
133+
**Video Tutorials:**
134+
- [LangChain Crash Course (YouTube)](https://www.youtube.com/watch?v=lG7Uxts9SXs)
135+
- [RAG Tutorial (YouTube)](https://www.youtube.com/watch?v=8OJC21T2SQ4)
136+
- [Building LLM Apps (YouTube)](https://www.youtube.com/playlist?list=PLIUOU7oqGTLieV9uTfD-7qHO8zJqkRnZC)
137+
- [Vector Databases Explained](https://www.youtube.com/watch?v=oZWVmJ5nP3U)
138+
139+
**Practice:**
140+
- [LangChain Templates](https://github.com/langchain-ai/langchain/tree/master/templates) - Example projects
141+
- [RAG Examples](https://github.com/langchain-ai/langchain/tree/master/templates/rag) - RAG implementations
142+
- [Hugging Face Spaces](https://huggingface.co/spaces) - Deploy and share GenAI apps
143+
- [LangChain Playground](https://smith.langchain.com/) - Experiment with LangChain
144+
145+
**[Complete Detailed Guide →](generative-ai-llms.md)**
146+
147+
**Additional Resources:**
148+
- [Generative AI Comprehensive Guide](../resources/generative_ai_comprehensive_guide.md) - Complete overview with all concepts
149+
- [RAG Comprehensive Guide](../resources/rag_comprehensive_guide.md) - Deep dive into RAG implementation
150+
- [LangChain Guide](../resources/langchain_guide.md) - LangChain framework details
151+
- [AI Agents Guide](../resources/ai_agents_guide.md) - AI agents and multi-agent systems
152+
- [GenAI Production Deployment](../resources/genai_production_deployment.md) - Production deployment patterns
166153

167154
---
168155

169-
**Remember**: Modern AI applications use pre-trained LLMs. This module teaches you how to build with them, not just train from scratch.
156+
**Previous Phase:** [12-natural-language-processing](../12-natural-language-processing/README.md)
157+
**Next Phase:** [19-sql-database-fundamentals](../19-sql-database-fundamentals/README.md)

0 commit comments

Comments
 (0)