Building AI solutions with open source

Chapters

• Architecture patterns for AI applications 

  Design patterns for integrating LLMs into applications. RAG, fine-tuning, prompt chaining, and agentic architectures.
• Choosing and evaluating open source models 

  Comparing open source LLMs, understanding model families, benchmarks, and selecting the right model for your use case.
• Infrastructure and deployment options 

  Self-hosting vs cloud APIs, GPU infrastructure, containerization, and deployment strategies for LLMs.
• Running LLMs locally and at scale 

  Setting up inference servers, load balancing, batching, and optimizing throughput for production workloads.
• Vector databases and embeddings 

  Implementing semantic search, building RAG systems, and choosing vector database solutions.
• Fine-tuning and customization 

  When and how to fine-tune models, dataset preparation, training infrastructure, and evaluating results.
• Cost optimization and monitoring 

  Tracking costs, optimizing inference, caching strategies, and monitoring model performance in production.
• Security and compliance 

  Data privacy, model security, prompt injection prevention, and compliance considerations for AI systems.
• Building an end-to-end AI application 

  Putting it all together: architecture, implementation, deployment, and maintenance of a production AI system.