Architect Expert

Architects to build the design and documentation scenarios frontier AI labs use to test model judgment.
AfterQuery builds the expert evaluation data that frontier AI labs use to measure and improve their models. We're hiring architects as contractors to author and grade the design and documentation scenarios those evaluations run on.

### What you'll actually do
- Author realistic architecture scenarios, from schematic design through construction documents
- Write reference answers with the design and code reasoning behind each decision
- Grade model output for design quality and code compliance
- Flag designs that read polished but would not survive plan review

### Who this is for
- Architects with a professional degree in architecture or an active license
- Practitioners with at least 2 years in design or construction documentation
- Architects with depth in building codes, construction documents, and design development
- Clear writers who can document the reasoning behind a design decision

### Contract and pay
- $100 to $125/hr, paid weekly via Stripe. Your rate is set from your experience, not negotiated down after you start.
- Fully remote, fully async. No fixed hours.
- Minimum 10 hours per week, no maximum. Scale up or down week to week.
- No screen recording, no keystroke monitoring, no idle timers.
- Independent contractor. Ongoing project work rather than single task batches.

### How our hiring process works
- Apply at /apply/architect-expert. About five minutes. Resume plus a few structured questions.
- Resume review. Every application gets read against a published rubric and gets a decision.
- Written assessment. You may be asked to take a written assessment for further placement.
- Onboarding. Sign the contractor agreement, connect Stripe, and get platform access.
- Project match. You're matched to live work in your specialty, usually within a week of onboarding.

### About AfterQuery
AfterQuery is an expert data lab. We build the evaluation and training data that frontier AI labs use to measure what their models can and can't do in real professional work. Our contributors are practicing professionals, not annotators, and the work is credited, reviewed, and paid at professional rates.