0.2 Who this guide is for (beginner → expert tracks)

Who this is for

This guide is for people who want to build real software faster by iterating with Google AI Studio, while keeping engineering control over architecture, correctness, and shipping.

You’ll get value if you’re any of the following:

• New to building software: you want a structured way to learn by shipping small projects.
• A working developer: you want faster scaffolding, refactors, debugging, and documentation without losing code quality.
• A product-minded builder: you want to prototype rapidly and then harden what works into something maintainable.
• A tech lead / founder: you want a practical workflow for using AI without turning your repo into spaghetti.
• A team: you want shared prompts, reviews, evals, and guardrails so AI-assisted work scales safely.

Who this is not for (yet)

This guide might be a frustrating starting point if:

• You want a “type one prompt, ship a product” promise.
• You don’t plan to run code locally or validate outputs.
• You’re trying to build a regulated production system and can’t invest in testing, security, and auditability.

Choose a track

Pick the track that matches your current baseline. You can always switch later.

• Beginner: new to coding or new to building projects end-to-end
• Intermediate: you can ship software, but want to use AI Studio effectively
• Expert: you care about scale: evaluation, RAG, agents, security, cost, and team workflows

Beginner track

Start here if: you’re learning fundamentals, you’ve never shipped a small app, or you mostly want a clear step-by-step path.

What you focus on

• Turning prompts into runnable projects quickly (and repeatedly).
• Asking for a plan before code, defining constraints, and writing “done” criteria.
• Debugging with minimal reproducible examples instead of long chats.
• Developing the habit: run → observe → refine.
• Basic tests or checks so you can change things without fear.

Recommended starting path

• Part I — Foundations
• Part II — Google AI Studio: First Contact
• Part III — Beginner Track

Intermediate track

Start here if: you can already build and ship software, but you want AI-assisted workflows that hold up under real constraints.

What you focus on

• Exporting prototypes into a repo with a clean architecture boundary around model calls.
• Structured output (JSON) and schema validation to reduce “parse vibes” failures.
• Tool/function calling patterns with error handling and budgets.
• Refactoring existing codebases safely (small diffs + tests).
• Lightweight evaluation habits to keep quality stable over time.

Recommended starting path

• Part II — Google AI Studio: First Contact
• Part V — Intermediate Track
• Part VI — Practical Workflows

Expert track

Start here if: you’re operating at production scale, building with sensitive data, or designing systems where reliability and security are as important as functionality.

What you focus on

• Evaluation harnesses, regression detection, and measurable quality bars.
• Reliability engineering: timeouts, retries, circuit breakers, caching, observability.
• Security and prompt-injection defense: threat modeling, least-privilege tools, secrets hygiene.
• RAG and grounding: retrieval pipelines, provenance, and “don’t make stuff up” design.
• Agentic workflows that stay auditable and under control.

Recommended starting path

• Part IX — Testing & Quality
• Part X — Security
• Part VIII — Retrieval & Grounding
• Part XIII — Expert Mode

Best way to jump in

If you’re unsure which track fits, use this quick self-check:

• If you want a guided path: start in Part III.
• If you want to turn prototypes into real repos: start in Part V.
• If you want to build production-grade AI features: start in Part IX and Part X.

Where to go next

• 0.3 How to follow along (projects, checkpoints, and “ship points”)
• 0.4 Required vs optional tools (AI Studio, editor, Git, runtime)
• Part I — Foundations