Leads the team this role joins
Heads the group this role sits in
In-house, owns the opening
From your own Gmail and LinkedIn.
About the role
This is a remote freelance role for a senior applied AI engineer who will build internal AI tools and an evaluation suite with a founder who wants to learn the system. The engagement runs one to three months at under 30 hours a week, and only individual freelancers are invited, not agencies. The client runs a small reverse-recruiting firm that applies to jobs on behalf of its own clients.
What you’ll do
- Build a role-fit classifier that scores job descriptions against a candidate profile and returns a structured verdict with reasoning.
- Build a grounded resume and narrative tool that drafts tailored bullets with strict citations.
- Build an outreach agent that researches companies and drafts messages, pausing for human approval before anything is sent.
- Set up an evaluation harness with at least 50 test cases covering grounding, classifier accuracy and an LLM-as-a-judge rubric.
- Pair with the client in live sessions of two to three hours a week and record your decisions in a GitHub repository.
What they’re looking for
- Production LLM features you have built yourself, not only work adjacent to a team that did.
- Focus on evaluation and failure modes, not just prompt tweaking.
- Ability to explain complex ideas in plain English before using technical terms.
- Working as an individual freelancer, not through an agency.
What’s on offer
- Less than 30 hours per week for one to three months.
- Remote, ongoing project.
Skills
Questions about this role
How many hours a week is this?
Less than 30 hours per week.
How long does it last?
One to three months.
Can agencies apply?
No. The posting is for individual freelancers only.
What should I send to apply?
A two-minute recorded walkthrough of a pipeline or repository you built, one thing you would change in the three-phase plan and why, and a real AI failure you caught and replaced.