“You haven’t shipped anything yet. That’s about to change.”
You’re raw. CS degree, bootcamp grad, self-taught — doesn’t matter. What matters is you’re hungry and you learn fast. You haven’t shipped professionally, and that’s fine. Every senior builder on this track once sat where you sit, staring at a codebase they didn’t fully understand.
This level is a Worca investment. You are not billing-ready. The company is betting that with the right mentorship and AI tools, you can become productive faster than anyone thought possible. Don’t waste that bet.
Your job: learn a stack with AI assistance, internalize git discipline, and ship something real in a guided environment. Not a tutorial project. Not a toy. Something that works, that someone else can read, that you’d put your name on.
What You Do
- Ship guided projects — complete structured assignments in a real codebase under mentor supervision
- Learn AI-assisted development — Claude Code, Cursor, and whatever tools your mentor prescribes. These are not optional. They are how you work.
- Git fundamentals — branching, committing, PRs, code review. Every change tracked. No cowboy commits.
- Read more code than you write — study the codebase you’re working in. Understand patterns before you add to them.
- Daily standups — what you did, what you’re doing, what’s blocking you. No hiding.
- Document what you learn — if you had to figure something out, write it down for the next person
AI Skills Required
- Use Claude Code or Cursor as your primary development environment — not occasionally, every session
- Generate boilerplate and scaffolding with AI, then understand what it produced
- Debug with AI assistance — describe the error, get suggestions, verify the fix yourself
- Read and summarize unfamiliar code using AI before modifying it
- Write commit messages and PR descriptions with AI drafting (human review before submitting)
- Use AI to learn new concepts — but verify against docs, don’t trust blindly
Self-Evaluation Checklist
Rate yourself honestly. If you can’t check every box, you’re not ready to level up.
- I can clone a repo, create a branch, make changes, and open a PR without hand-holding
- I use AI tools in every coding session — they’re habit, not novelty
- I’ve shipped at least one feature that passed code review and merged to main
- I can read a function I didn’t write and explain what it does within 5 minutes (with AI assist)
- I write clean commit messages that explain why, not just what
- I ask questions when stuck instead of spinning for hours
- I can set up a local dev environment from a README without someone walking me through it
- My mentor doesn’t have to remind me about the same mistake twice
Training Curriculum
Month 1: AI Tooling and Dev Environment
- AI Dev Tools Bootcamp — hands-on with Claude Code, Cursor, GitHub Copilot. Learn when each tool shines and when it misleads.
- Git from Scratch — branching strategies, rebasing vs. merging, writing good PRs, reviewing others’ code
- Environment Setup — Docker, package managers, local dev servers. You should be able to spin up any project in the track’s standard stacks.
- Prompt Engineering for Code — how to ask AI for what you actually need. Specificity. Context. Iteration.
Month 2: Guided Building
- First Feature Assignment — a real, scoped task in a real codebase. Mentor-defined spec. You build it.
- Code Reading Practice — assigned codebase walkthroughs. Read 500+ lines of production code per week. Summarize what you find.
- Testing Basics — write tests for your own code. AI can help generate test cases, but you need to understand what’s being tested.
- Pair Programming — weekly sessions with your mentor or an L3+. Watch how they think, not just what they type.
Month 3: Ship and Assess
- Second Feature Assignment — slightly less guidance this time. Spec is clear, but you plan the implementation.
- Code Review Practice — review PRs from other L1s. Give feedback. Learn to spot issues.
- Portfolio Assembly — document what you built, what you learned, how you used AI. This is your L2 application.
- Mock Assessment — practice the L2 qualification with your mentor. Get honest feedback.
Ranking Standard
| Metric | Threshold | How It’s Measured |
|---|---|---|
| Features shipped | 2+ merged PRs | Git history |
| AI tool usage | Every coding session | Mentor observation + workflow log |
| Code review pass rate | 80%+ first-submission approval | PR history |
| Git hygiene | Clean branches, clear commits | Repo audit |
| Learning velocity | Measurable week-over-week improvement | Mentor assessment |
| Blocker communication | Flags within 1 hour, not end of day | Standup log + Slack history |
Promotion to L2
Requirements
- Minimum 3 months at L1
- Pass L2 qualification assessment:
- Live coding exercise — given a clear spec, build a small feature in 90 minutes using AI tools. Panel observes your process, not just the output.
- Code review — review a PR with intentional issues. Identify what’s wrong and explain why.
- AI workflow demonstration — show the panel your actual dev workflow. How do you use AI? Where do you not?
- Portfolio review — present the features you shipped. Walk through the code. Explain your decisions.
- Positive mentor feedback
- Zero instances of going dark on blockers
What the Panel Looks For
- Learning speed — are they measurably better than month 1? The slope matters more than the intercept.
- AI fluency — is AI integrated into how they think about problems, or is it an afterthought?
- Shipping discipline — do they finish things? Do PRs sit open for weeks, or do they drive to completion?
- Coachability — do they take feedback and apply it, or do they nod and repeat the same mistakes?
Mentorship at This Level
- You receive: An L4+ mentor assigned to you. Weekly 1:1 check-ins (45 min). Your mentor reviews your code, assigns projects, and decides when you’re ready for L2.
- You give: Nothing yet. Your only job is to learn and ship. Mentorship responsibilities start at L4.
- Pair sessions: At least 2 pair programming sessions per month with L3+ builders. Observe their workflow. Ask questions after.
What You Don’t Get Yet
- Client-facing work (you’re not billing-ready)
- Autonomy on scope or architecture decisions
- Mentee slots
- Choice of stack (you work in what your mentor assigns)
These unlock as you prove yourself. Earn them.