AI_devs 4 Builders

AI Engineering Course, Building Production-Ready AI Solutions

Completed a 5-week cohort-based course focused on building production-ready AI solutions. The program consisted of 25 lessons, 25+ hands-on projects, 5 live meetings, ongoing instructor support, and collaborative discussions with other participants.

Areas of Knowledge

Using Large Language Models in Code

• Working with Large Language Models through APIs
• Steering LLM behavior through prompts and code
• Structuring LLM responses using JSON Schema
• Using Function Calling and native API tools
• Designing high-quality tools for agents (CLI and MCP)
• Processing multimodal content: image, audio, and video
• Managing limitations and security concerns such as prompt injection
• Optimizing cost, performance, and reliability

Context Engineering

• Understanding the difference between Prompt Engineering and Context Engineering
• Deep understanding of the role of context in steering LLMs
• Building interactions that prioritize prompt cache
• Building tool outputs that steer agent behavior in a generalized way
• Balancing static and dynamic information without losing prompt cache
• Practical use of context compression and extraction techniques
• Context Engineering in multi-agent architectures
• Designing high-quality instructions, including meta-prompts
• Using the non-deterministic nature of LLMs as an advantage

Observability and Evals

• Understanding the role of observability and the value of evals
• Designing architecture with observability and evals in mind
• Designing both offline and online evals for prompts and agents
• Designing datasets and sourcing examples from production
• Evals for single- and multi-turn interactions
• Using evals for filtering and moderating unwanted behaviors
• Using evals to increase reliability and performance of the entire system
• Building eval-driven integrations and agentic tools

Building Production Apps

• Designing, maintaining, and scaling generative app architecture
• Understanding new challenges driven by the fast pace of AI progress
• Building multi-agent systems that operate in the background
• Building tools that work beyond the chat interface and integrate with existing logic
• Designing systems that optimize business processes with human-in-the-loop
• Using modern tooling with informed decisions about the tech stack
• Understanding the challenges of AI frameworks and modern tools
• Building AI solutions that match business requirements with AI capabilities
• Addressing challenges and opportunities of generative apps

Course Instructors

• Adam Gospodarczyk
• Jakub Mrugalski
• Mateusz Chrobok