Course overview
What your team will get out of this course
Automate documentation updates with agents
Use AI for coding suggestions with open source LLMs
Master agent-led code reviews and quality checks
Gain advanced debugging skills with agents
Generate usable code by instructing agents
How the course is structured
- The course has 5 sprints, each two weeks long.
- In every sprint, there's live instruction on Zoom and a guided project.
- You'll build a fully-functional agent in each sprint.
A week in the life
Designed for busy professionals, our course includes two live 1-hour classes per week with an optional office hours every week.
Weekly Schedule
- 2 Live Classes: 1-1.5 hrs each
- Office Hours: 1-1.5 hours (optional)
- Outside Class: 2-3 hours on projects (recommended)
This course includes
Interactive live lessons online
Guided feedback and reflection
5 AI projects to apply learnings
Direct access to an instructor
Private community of peers
Certificate upon completion
What your team will learn
Tracing and LLM Essentails
OpenAI ◦ Ollama ◦ LangSmith ◦ LangFuse
- LangSmith: Tracing outputs and annotating data.
- Choosing the right LLM: Guide to selecting suitable LLMs.
- Prompt engineering: Strategies for effective prompt design.
Retrieval-Augmented Generation
LangChain ◦ Pinecone ◦ Chroma ◦ Nomic
- Fundamental chaining: Basic chaining techniques and strategies.
- Vector embeddings: Enhancing AI with vector embeddings.
- RAG: Combining retrieval with generative models.
Designing and Building Agents
LangGraph ◦ OpenAI ◦ Docker
- Agent Strategy: How to design and build agents
- Agents vs. Chains vs. Graphs: Which tool for each specific tash
- Agent-Supervisor: Complex frameworks for building agents
Deploying Agents
LangServe ◦ Docker ◦ OpenAI
- Building performative AI apps: Creating production AI apps
- Connecting to workflows: Integrating AI with company workflows
- Advanced chaining: Complex chaining techniques for scalability
Fine-tuning
HuggingFace ◦ PEFT ◦ QLoRA ◦ OpenAI
- How to fine-tune OpenAI: Considerations and strategy
- Advanced techniques: PeFT, LoRA, and QLoRA
- Testing strategy: Fine-tuning base models on HuggingFace
Multi-Agent Systems
LangGraph ◦ CrewAI
- Multi-Agent Systems: Solving problems with multi-agent systems.
- Agent decision making: Decision making with LangGraph.
- Designing solutions: Considerations for building prototypes.
Capstone
LangGraph ◦ Ollama ◦ OpenAI ◦ LangChain
- Complex prototype: Implement solution inside your business ecosystem
- Expert guidance: Staff designs and builds your idea with you
- Viable projects ready for iteration: Build your idea with our team
Build agents in every project
What are AI agents
- Intelligent software programs designed to perform tasks autonomously.
- They make decisions and execute actions with minimal intervention.
- Agents continually adapt and learn to optimize their performance.
Tech Writer
Coder
QA Tester
Architect
Who is this course for
Software engineers and developers
Engineering leaders and managers
Data scientists and ML engineers
Why build agents over using generic tools
There's an AI gold rush unfolding, and it's essential for every business to develop custom solutions to stay ahead.
Where you are
- Generating code with GitHub copilot
- Creating simple OpenAI API wrappers
- Processing documents with ChatGPT
- Building prompt libraries
What we'll teach you
- Developing code review agents
- Automating technical doc updates
- Implementing accurate code generation
- Building multi-agent systems
Unlike generic off-the-shelf tools, custom AI solutions are tailored to your unique needs, giving you full ownership.
Custom AI apps are
✅ Tailored to your specific needs
✅ Made to seamlessly integrate in your ecosystem
✅ Designed to give you a competitive advantage
✅ Built with scalability and flexibility in mind
✅ Cost-effective compared to generic solutions
✅ Meant to give you full ownership and control