I finished Python Beginner. Maybe Automation too. What's different about this track?
One new layer: calling LLMs — large language models — from Python. Send a prompt, get a response. Same shape as a tool call (one function call, one return value), but the response is generated text, not a deterministic API result.
Why does that matter?
Three new things to learn. Prompting — the input is plain English; small wording changes can swing the output. Structured output — getting JSON back when you want JSON, and recovering when the model returns prose around it. Multi-turn — threading context across a conversation so the model remembers what was said.
And what's the same?
Everything else. The Python primitives from Beginner — variables, lists, dicts, functions, try/except — all apply. The patterns from Automation Beginner if you've done it — helpers, retry, batch — work identically with LLM calls. By week 4 you'll write batch LLM scripts that classify a list of items, retry on bad output, and track cost.
AI Beginner adds the third primitive layer: language models.
Week 1 — Mental model + calling. What an LLM technically is, calling one with Agent(model).run_sync(prompt), prompt specificity, iterating on prompts, the four task verbs (summarise, infer).
Week 2 — Output shape. Transform, expand, few-shot examples, structured JSON output, handling parse failures, schemas in prompts.
Week 3 — Multi-turn. Conversations as message lists, system prompts, token cost awareness, refusals, a synthesis composing 5 prior primitives.
Week 4 — Production patterns. Batch classification, LLM helpers, retry on parse failure, self-consistency, cost-aware batching, final integration.
Write small Python scripts that call an LLM, parse structured output reliably, hold a multi-turn conversation, and run batched calls with retry and cost tracking. That foundation lets you build classifiers, extractors, summarisers, and chatbots — composing what you learned in Python Beginner with the new LLM layer.
Python Beginner (or equivalent fluency). Comfortable with variables, lists, dicts, conditionals, for loops, functions, try/except, and JSON parsing. Auto Beginner is helpful but not required — week 4 production patterns parallel Auto-Beg's but stand alone here.
Rate each statement honestly on the 1-5 scale. The same prompts come back on day 30 to mark your delta.
I finished Python Beginner. Maybe Automation too. What's different about this track?
One new layer: calling LLMs — large language models — from Python. Send a prompt, get a response. Same shape as a tool call (one function call, one return value), but the response is generated text, not a deterministic API result.
Why does that matter?
Three new things to learn. Prompting — the input is plain English; small wording changes can swing the output. Structured output — getting JSON back when you want JSON, and recovering when the model returns prose around it. Multi-turn — threading context across a conversation so the model remembers what was said.
And what's the same?
Everything else. The Python primitives from Beginner — variables, lists, dicts, functions, try/except — all apply. The patterns from Automation Beginner if you've done it — helpers, retry, batch — work identically with LLM calls. By week 4 you'll write batch LLM scripts that classify a list of items, retry on bad output, and track cost.
AI Beginner adds the third primitive layer: language models.
Week 1 — Mental model + calling. What an LLM technically is, calling one with Agent(model).run_sync(prompt), prompt specificity, iterating on prompts, the four task verbs (summarise, infer).
Week 2 — Output shape. Transform, expand, few-shot examples, structured JSON output, handling parse failures, schemas in prompts.
Week 3 — Multi-turn. Conversations as message lists, system prompts, token cost awareness, refusals, a synthesis composing 5 prior primitives.
Week 4 — Production patterns. Batch classification, LLM helpers, retry on parse failure, self-consistency, cost-aware batching, final integration.
Write small Python scripts that call an LLM, parse structured output reliably, hold a multi-turn conversation, and run batched calls with retry and cost tracking. That foundation lets you build classifiers, extractors, summarisers, and chatbots — composing what you learned in Python Beginner with the new LLM layer.
Python Beginner (or equivalent fluency). Comfortable with variables, lists, dicts, conditionals, for loops, functions, try/except, and JSON parsing. Auto Beginner is helpful but not required — week 4 production patterns parallel Auto-Beg's but stand alone here.
Rate each statement honestly on the 1-5 scale. The same prompts come back on day 30 to mark your delta.
Create a free account to get started. Paid plans unlock all tracks.