Course Outline
Course Outline Training Proposal
Day 1 - Foundations of AI and Python for Data Workflows
• Survey of the current artificial intelligence and machine learning landscape
• The impact of AI on modern data engineering practices
• Refresher on Python essentials for AI applications
• Data manipulation using pandas and NumPy
• Basics of API interaction and JSON data processing
• Practical exercise: loading and transforming datasets
Day 2 - Machine Learning Essentials for Practitioners
• Principles of supervised and unsupervised learning
• Techniques for feature engineering and data preparation
• Fundamentals of model training with scikit-learn
• Assessing model evaluation and performance indicators
• Overview of model deployment methodologies
• Hands-on session: building a basic predictive model
Day 3 - Introduction to LLMs and Prompt Engineering
• Gaining insight into large language models and their operational mechanics
• Exploring tokenization, context windows, and inherent constraints
• Core principles and methods for prompt design
• Application of zero-shot and few-shot prompting
• Strategies for prompt evaluation and iterative refinement
• Practical prompt engineering activities
Day 4- Building AI Applications with LLMs
• Utilizing LLM APIs within Python environments
• Concepts of structured outputs and function calling
• Developing chat-based and task-oriented applications
• Introduction to retrieval-augmented generation
• Linking LLMs with external data sources
• Mini-project: creating a basic AI assistant
Day 5 - Operationalizing AI Solutions
• Architecting scalable AI workflows
• Embedding AI into data pipelines
• Monitoring and enhancing model performance
• Strategies for cost optimization and API management
• Considerations regarding security and responsible AI
• Capstone project: developing an end-to-end AI solution
Testimonials (2)
The trainer was very available to answer all te kind of question I did
Caterina - Stamtech
Course - Developing APIs with Python and FastAPI
Trainer develops training based on participant's pace
