PixioLab
Cover of Guide 00, The 45-Day AI Engineer Roadmap

Guide 00 · 29 pages · 11 chapters · free

The 45-Day AI Engineer Roadmap

Start here. The week-by-week plan that ties every other guide together.

By PixioLab

Free with an account. It's also the plan the 45-day path follows. See the 45-day roadmap.

What's inside

  1. 01

    Plan, rules and costs

    Before Day 1, pick a track, check the prerequisites, decide what you can compress, set up your machine, and know what this will cost. Most of it is free, but calling a model costs a little.

    Two tracks, one destination · Prerequisites, and what you can compress · Day 1 setup · Nine rules of the road · What's free, what's freemium, what costs money · A realistic API budget

  2. 02

    The capstone: one repo, built in layers

    You build one system across the 45 days, adding a layer each week. By Day 42 it is feature-complete, so the last three days go on polish and publishing rather than panic.

    What you're building · How the repo grows · Repo layout

  3. 03

    How LLMs work, and your first call

    AI engineers build products on top of pretrained models rather than training them. Start with a working mental model of what a model does, then make your first calls from code and record what the numbers tell you.

    Concepts · Resources · Build: your first logged calls

  4. 04

    Prompts, schemas and tools

    Get fluent at the main interface of the job: clear prompts, outputs checked against a schema, tools the model can ask you to run, and calls that survive timeouts and rate limits. By Sunday you'll also have chosen your capstone corpus.

    Concepts · Resources · Build: the extractor CLI, and your corpus

  5. 05

    Retrieval and RAG: Sourcebook v0.3

    Give a model access to your corpus: ingest, chunk, embed and index it, retrieve with hybrid search, and answer with citations that point to real passages. This week builds the core of the capstone.

    Concepts · Resources · Build: Sourcebook v0.3

  6. 06

    Agents and tools: a bounded loop

    Move from single calls to multi-step systems, and learn when not to. Wrap Sourcebook's retriever as a tool, build a tool-use loop with hard budgets and a trace, and measure whether the agent actually beats plain RAG on your questions.

    Concepts · Resources · Build: Sourcebook agent mode

  7. 07

    Evaluation and production: Groundtruth

    Prove Sourcebook works with numbers, then make it observable, reliable and cheaper. Build Groundtruth, a retrieval evaluation harness, add a checked judge for answer quality, and log cost and latency per request. This is the week that turns a demo into evidence.

    Concepts · Resources · Build: Groundtruth v1

  8. 08

    Safety, ML foundations and the ship gate

    Make Sourcebook resist the obvious attacks and leaks, wire up Tripwire so a regression can't merge, and learn just enough classic ML for interviews. Day 42 is a buffer, and by its end the capstone is feature-complete.

    Concepts · Resources · Build: safety layer, Tripwire and an ML notebook

  9. 09

    Polish, publish, apply

    Because the capstone grew week by week, these three days add no features. You write it up, put it where people can see it, and start applying. If you're behind, follow the triage rule below rather than squeezing in more.

    Resources

  10. 10

    Get hired

    Your repo is the proof. This chapter turns it into a README people read, resume lines that survive follow-up questions, and a four-week interview and job-search loop through Guides 06–09.

    The portfolio README · Talk about the project in two minutes · The interview path through Guides 06–09 · After Day 45: what to build next

  11. 11

    Resource index

    All 69 resources, grouped by the week that uses them. Every link was opened and checked in November 2026; where an address redirects, the current one is printed.