Skip to main content
Skip to content
Menu

AI pretraining Beginner 3 parts

Write a decision brief comparing the speed-vs-quality tradeoff for your coding workflow

You'll have a 1-page analysis comparing model options for your specific coding task (speed for quick tasks vs. reasoning for complex problems), with cost-per-task estimates and a recommendation tied t

What you'll make

Coding Model Decision Brief

  1. Current Usage: your baseline coding tasks and current model choices
  2. Speed vs Quality Matrix: measured performance on your actual code types
  3. Economic Breakpoints: cost-per-task analysis and switching thresholds
  4. Workflow Recommendation: which model for which coding scenario

Write a structured analysis with specific cost numbers and task examples from your own testing

The 3 parts

Each part ends with a check you can see.

  1. Map Your Coding Reality

    Document what coding tasks you actually do and how much time you spend on different types of problems

    Done when

    You have 10-15 recent coding tasks categorized by complexity with time estimates for each type

  2. Run Head-to-Head Model Tests

    Test 2-3 different AI models on the same coding tasks to measure actual speed and quality differences

    Done when

    You have timed comparisons of at least 2 models on 3-5 coding tasks, with quality scores based on how much editing each output required

  3. Design Your Model Switching Strategy

    Create a decision framework that determines which model to use based on task characteristics you can identify upfront

    Done when

    You have a workflow decision tree with specific trigger conditions for switching models, tested against 5 real coding scenarios from your work

    Capstone: you put together your Coding Model Decision Brief.

Where this helps

An elective: a single lesson that sits outside rtwo's job certificates and doesn't count toward one.

About this lesson

What you'll be able to do

  • Compare inference speed and quality trade-offs for your actual coding tasks using real model data
  • Calculate cost-per-task estimates for different models based on your usage patterns
  • Make an evidence-backed recommendation for which models to use when in your development workflow
  • Identify the specific quality thresholds where switching models makes economic sense

Who it's for

  • Developers who use AI coding assistants and want to optimize their model choices
  • Engineers curious about the economics behind AI model selection
  • Anyone spending significant time or money on AI coding tools who wants data-driven usage decisions

What it won't cover

  • Training your own models or fine-tuning
  • Deep technical implementation of MoE or distributed training
  • Comprehensive model benchmarking across all possible coding tasks
  • Setting up inference infrastructure from scratch

Skills you'll practise

  • Cost-benefit analysis for AI model selection
  • Quality threshold identification through direct testing
  • Workflow optimization based on task complexity
  • Economic reasoning about AI tool usage