The road from "I like art and I'm not scared of code" to a technical artist job - laid out in order, honestly, with every step pointing at the exact page to read next. Nobody follows a roadmap perfectly. That's fine. It exists so you always know what's roughly next.
Tech art is for people who get equal satisfaction from a beautiful render and a script that saves someone an hour a day. You don't need to be a math genius, a great artist, or a great programmer - you need to be adaptable, curious, and genuinely happy helping other people work faster. Read the role page, then the specializations map, and see if anything in there makes you lean forward.
Python is the TA's multitool - learn it once, use it in every DCC you'll ever touch. You don't need computer-science depth; you need files, paths, strings, dicts, loops, and functions to feel boring. The finish line for this stage isn't a certificate - it's shipping your first pipeline script and feeling the "wait, that just saved me twenty minutes" click.
Choose one 3D package and learn to script it. Blender is free and where most people should start; Maya still runs most AAA character pipelines; Houdini is the deep end with the biggest payoff. Don't try to learn all three - depth in one beats a survey of all. While you're here, learn version control. It's a day of effort and it's on every job description you'll ever read.
Studios don't hire TAs who can't get an asset into an engine and explain why it runs slow. Learn Unreal's import pipeline, make some materials, and read the performance section until "draw call" and "overdraw" are words you use naturally. If your target studios use Unity, skim the Unity page too. The concepts transfer almost one-to-one.
Pick the specialization that made you lean forward back in week zero and go deep: shaders, procedural, pipeline, rigging, any of them. Then build the portfolio formula: one tool, one shader, one pipeline. Three pieces with honest breakdowns beat thirty renders. The project briefs give you real tickets with constraints and a definition of done, so you are not staring at a blank page wondering what to build.
Treat the job hunt as a pipeline problem - inputs, process, iteration. A one-page resume that leads with what you've built, a portfolio a lead can evaluate in ninety seconds, and interview prep that's about reasoning patterns rather than memorized answers. Expect rejection; every TA you admire has a folder of them.
The first 90 days of a first TA job are their own skill. Learn the studio's pipeline before proposing to fix it, make friends with one artist and one engineer, and write down every acronym you don't know. Then look it up in the glossary. That is exactly why it exists. From there the job compounds: every tool you ship makes the next one easier to justify.
The best junior TA I ever hired had no degree and a portfolio of exactly three things: a Blender add-on that renamed things against a real studio convention they had found in a GDC talk, a dissolve shader with a two-page breakdown of every mistake they made building it, and a Python script with actual tests. We interviewed eleven people with prettier reels. We hired the one who showed their work.
Breakdowns beat beauty renders. The one-tool-one-shader-one-pipeline formula, what leads actually look at, and how to present code without apologizing for it.
-> CareerOne page. Lead with what you've built. Survive the keyword filter without writing like a robot. Includes a copy-paste template.
-> CareerHow TA interviews actually run, plus a bank of real questions - Python, math, shaders, pipeline, behavioral - each with how a strong answer reasons.
-> CareerDirectional numbers by region and level, what moves them, and word-for-word scripts for the conversation everyone dreads.
-> CareerThe first 90 days, decoding studio jargon, speaking artist and engineer, and protecting yourself from crunch without torching relationships.
-> CareerPerforce and Git for people who make binary files. Daily survival commands, golden rules, and how to fix the classic disasters.
->