Visão Geral

Este curso apresenta o desenvolvimento de aplicações de Inteligência Artificial Generativa utilizando modelos open source. O participante aprenderá a trabalhar com Large Language Models (LLMs) de código aberto, implementar soluções locais e em nuvem, construir arquiteturas RAG (Retrieval-Augmented Generation), desenvolver agentes inteligentes e integrar modelos generativos em ambientes corporativos. O curso explora o ecossistema open source de IA, incluindo implantação, customização, otimização e governança de modelos.

Objetivo

Após realizar este curso, você será capaz de:

• Compreender o ecossistema de modelos open source para IA Generativa
• Implantar e administrar modelos generativos em ambientes locais e cloud
• Desenvolver aplicações utilizando LLMs open source
• Implementar arquiteturas RAG para acesso a conhecimento corporativo
• Construir agentes inteligentes e fluxos de automação baseados em IA
• Aplicar práticas de segurança, observabilidade e governança em ambientes de IA Generativa

Publico Alvo

• Desenvolvedores de software
• Engenheiros de IA e Machine Learning
• Cientistas de dados
• Arquitetos de soluções
• Engenheiros de plataforma e DevOps
• Profissionais interessados em soluções de IA Generativa open source

Pre-Requisitos

• Conhecimentos de programação (preferencialmente Python)
• Familiaridade com APIs, bancos de dados e aplicações web
• Conhecimentos básicos de Machine Learning e Inteligência Artificial
• Noções de Linux e containers são recomendadas

Conteúdo Programatico

Module 1: Introduction to Open Source Generative AI

1. Overview of Generative AI ecosystem
2. Open source versus proprietary models
3. Benefits and challenges of open source AI
4. Popular open source LLMs
5. Enterprise adoption scenarios
6. AI application development lifecycle

Module 2: Open Source Foundation Models

1. Large Language Models fundamentals
2. Transformer architecture overview
3. Model capabilities and limitations
4. Context windows and tokenization
5. Model selection criteria
6. Evaluating open source models

Module 3: Open Source AI Platforms and Ecosystem

1. Model repositories and communities
2. Local and cloud deployment options
3. AI development frameworks
4. Model serving concepts
5. Infrastructure requirements
6. Enterprise architecture considerations

Module 4: Running and Managing Local Models

1. Local inference fundamentals
2. Hardware requirements and optimization
3. GPU and CPU considerations
4. Model quantization concepts
5. Performance tuning techniques
6. Operational best practices

Module 5: Prompt Engineering for Open Source Models

1. Prompt design fundamentals
2. Zero-shot and few-shot prompting
3. Context management techniques
4. Structured output generation
5. Prompt evaluation methods
6. Prompt optimization strategies

Module 6: Building Generative AI Applications

1. Application architecture patterns
2. Conversational AI solutions
3. Knowledge assistants
4. Content generation systems
5. Workflow automation applications
6. User experience considerations

Module 7: Embeddings and Vector Databases

1. Embedding models overview
2. Semantic search fundamentals
3. Vector databases concepts
4. Similarity search techniques
5. Knowledge retrieval strategies
6. Enterprise search implementations

Module 8: Retrieval-Augmented Generation (RAG)

1. RAG architecture fundamentals
2. Document ingestion pipelines
3. Chunking and indexing strategies
4. Retrieval optimization techniques
5. Context enrichment approaches
6. Enterprise RAG implementations

Module 9: Fine-Tuning and Model Customization

1. Fine-tuning fundamentals
2. Instruction tuning concepts
3. Domain adaptation techniques
4. Parameter-efficient tuning methods
5. Dataset preparation strategies
6. Model evaluation and validation

Module 10: AI Agents and Autonomous Systems

1. Agent architecture concepts
2. Tool calling and orchestration
3. Multi-agent systems overview
4. Autonomous workflows
5. Enterprise automation scenarios
6. Agent governance considerations

Module 11: Security, Governance and Operations

1. AI security fundamentals
2. Data privacy and protection
3. Model governance practices
4. Observability and monitoring
5. LLMOps concepts
6. Compliance and risk management

Module 12: Capstone Project and Enterprise Implementation

1. End-to-end AI application development
2. Open source RAG solution project
3. AI agent implementation workshop
4. Model deployment exercises
5. Security and governance validation
6. Final enterprise Generative AI solution project using open source models