AI-Native Full-Stack Developer

AI-Native Full-Stack Developer

Lacroo Technologies Pty Ltd

Job description

About the role

We are looking for an exceptional AI-native full-stack product engineer who can take product specs, prototypes, or PRDs and turn them into clean, tested, QA’d, merge-ready branches. We are building a fast-moving software company and currently use AI coding tools heavily in our development workflow. Right now, the delivery process is: prototype review, spec clarification, implementation planning, AI-assisted coding, QA, checking what was actually built, and preparing features for final merge. We want to have this entire workflow managed by the right engineer. Your job will be to take a feature brief, PRD, prototype, or Loom walkthrough and own the process from: Input → clarification → implementation plan → AI-assisted build → self-review → QA → PR → merge-ready branch. We are looking for someone who can take ownership, not create more management work.

What You’ll Do

You will be responsible for:

• Reviewing PRDs, prototypes, tickets, or Loom walkthroughs

• Asking clear, batched clarification questions

• Creating concise implementation plans before coding

• Using AI coding tools to accelerate development

• Reviewing and improving AI-generated code

• Building full-stack product features

• Writing or updating tests where appropriate

• Running QA before handing work over

• Creating clean, reviewable pull requests

• Documenting what was built, what changed, and what risks remain

• Preparing branches that are actually ready to merge

• Communicating clearly and asynchronously

• What “Done” Means

A feature is not done when the code is written. A feature is done when:

• The implementation matches the PRD/prototype

• Acceptance criteria are completed

• Code is clean and understandable

• Build/lint/type checks pass

• Relevant tests are added or updated

• Manual QA has been completed

• Screenshots or Loom walkthrough are provided when relevant

• PR description is clear

• Risks, assumptions, and deviations are documented

• The branch is ready for review/merge

• Ideal Candidate

You are probably a strong fit if you are:

• An experienced full-stack engineer with strong product judgment
• Experienced with AI coding tools such as Claude Code
• Comfortable taking a vague product idea and turning it into a working feature
• Able to review AI-generated code critically
• Strong with GitHub PR workflows
• Detail-oriented with good QA discipline
• Clear in written communication
• Comfortable working in a fast-moving startup environment
• Able to work independently without needing constant hand-holding
• Good at asking the right questions early

You Are Not a Fit, please do not apply if:

• You blindly trust AI-generated code

• You need every task broken down into tiny instructions

• You say “done” without testing your work

• You create large messy PRs

• You cannot explain your own code

• You are only comfortable with narrow frontend or backend tasks

• You are not comfortable with ambiguity

• You dislike writing implementation plans or QA notes

• You require constant synchronous direction

• Required Skills

Please include your experience with:

• Full-stack web development

• Frontend frameworks

• Backend/API development

• Git/GitHub

• Pull request workflow

• Testing and QA

• AI-assisted coding tools

• Debugging and self-review

• Product-focused software development

• Nice to Have

• Startup experience

• Experience as a founding engineer or product engineer

• Strong UX/product taste

• Experience with automated testing

• Experience with CI/CD

• Experience with preview deployments

• Experience turning Figma/prototypes into production features

• Experience working directly with founders

• How We Work

We will usually provide one or more of:

• PRD
• Prototype
• Figma
• Loom walkthrough
• Linear/Jira/Notion ticket
• Product brief
• Screenshots
• Existing codebase

You will be expected to return:

• Clarifying questions, if needed

• Implementation plan

• Branch/PR

• QA notes

• Test results

• Screenshots/Loom if relevant

• Clear summary of what was built

• Any risks, tradeoffs, or unresolved questions

• What Success Looks Like

• Success means the team can give you a PRD or prototype and receive back a merge-ready branch without needing to manage every step.

Great output looks like:

• “I reviewed the feature brief, found two ambiguities, made reasonable assumptions, implemented the core flow, added tests, QA’d happy path and edge cases, documented one tradeoff, and the branch is ready for review. Management's input is only needed on this one product decision.”

• Preference will be given to candidates who can work with high ownership and potentially become a long-term product engineering partner.