Live online
Advanced Prompting and Evaluation — 1–2 Dec 2026
Two days on prompt architecture and measurement: decomposition, structured output, grounding, golden sets, model-graded scoring and regression testing.
When
1–2 Dec 2026
09:00–16:00 +0530 · 03:30–10:30 UTC
| When | How |
|---|---|
| Tuesday 1 Dec 2026 | 09:00–16:00 |
| Wednesday 2 Dec 2026 | 09:00–16:00 |
Where
Online
Asia/Colombo
About this course
Advanced
2 days · 12 hours
Two days on prompt architecture and measurement: decomposition, structured output, grounding, golden sets, model-graded scoring and regression testing.
What you'll be able to do
- Decompose an overloaded prompt into a chain of small steps you can test one at a time.
- Separate instruction, context and data so a prompt can be edited without being rewritten.
- Select few-shot examples that correct a specific observed failure rather than padding the context.
- Force output against a JSON schema and validate it before anything downstream receives it.
- Write a tool description precise enough that the model calls it with the right arguments.
- Ground an answer on your own documents, with citations and an honest refusal when the source is silent.
What's included
- Class recording, available for 12 months
- Evaluation harness: spreadsheet template and an optional Python notebook
- Prompt pattern library built across the two days, organised by failure type
- Rubric, golden-set and regression-log templates you can reuse on the next task
- MyLearnPlus certificate of completion
- Instructor email support after the class for questions on your own evaluation set
- One free repeat of the same live class within 12 months, on the terms shown on the booking page
