Building with large language models
For people who can write some code and want to build on top of models rather than just use them. Covers API calls, structured output, retrieval-augmented generation, evaluation and the cost and safety decisions that come with deploying something real.
6 modules · 27 lessons · taught with Python, OpenAI API, Anthropic API, Supabase
About 10 hours: 8 hours 55 minutes of lessons, including their exercises, plus 60 minutes for the assessment.
What you will be able to do
- Call OpenAI and Anthropic APIs from Python with structured output
- Build retrieval over a folder of PDFs
- Write an evaluation set and measure quality honestly
- Deploy a small assistant behind a web form
On completion
Codality verified certificate: Building with LLMs, issued to your profile and shareable with one link.
Modules
- 01Free
APIs, tokens and cost
4 lessons · 70 min
- 02Plus
Structured output and tool use
5 lessons · 95 min
- 03Plus
Retrieval over your own data
6 lessons · 120 min
- 04Plus
Evaluation and safety
5 lessons · 100 min
- 05Plus
Shipping the assistant
6 lessons · 150 min
- 06Plus
Assessment and project review
1 lesson · 60 min
Also on the AI track