AI Agent Tester - Real-World Tasks

$25–45/hr

rate as listed

Workload: ~15h/week
Engagement: Project-based
Location: US
Field: Generalist

Test two AI agents on everyday errands such as bookings, purchases, returns and forms, then compare their performance using recorded evidence. The work is U.S.-based, completed from a smartphone and requires personal account access, careful step-by-step documentation and written scoring.

What the work involves

You will run the same kinds of real-world tasks through two AI agents, including bill payments, shopping, groceries, travel bookings, returns, email follow-ups, family logistics and forms. For each attempt, you will approve or stop important actions, screen record the full run and document what happened.

How runs are evaluated

Each run is scored for completion, output quality, control issues, user satisfaction and time. The platform is likely to screen for whether you can follow a test protocol, distinguish successful completion from a merely plausible attempt, identify unsafe or unwanted agent behavior and support your ratings with concrete evidence. Clear written feedback matters more than coding knowledge, and no coding or AI background is required.

Practical requirements

This is a U.S.-based contractor role completed remotely and primarily from your phone, on an asynchronous schedule. A run takes about 1.5 hours including evidence collection and scoring. You need an active personal AI assistant account on a Plus tier, active Muse and Dots access, a smartphone capable of screen recording and a payment card for test purchases that are reimbursed afterward.

Timing and compensation

The project is already active. The listing says selected contractors may receive an offer and complete onboarding within 24 to 48 hours, with work beginning when tasks are assigned, so task volume is not stated or guaranteed. The observed compensation range is:

Pay band: $25–45/hr

What they look for

ChatGPT
Muse
AI agents
Attention to Detail
Following Instructions
Screen Recording

Included with this listing

Interview prep — 10 questions

Matched to what this platform's screening actually measures. Answer each out loud in 60–90 seconds: claim → concrete example → implication.

What it tests: This tests whether your experience includes concrete, multi-step work comparable to bookings, purchases, forms, returns or office operations rather than only general interest in AI.
A strong answer: Give specific roles, dates or durations, and examples of workflows you personally completed or monitored. Clarify where you had to follow instructions, verify outcomes and document exceptions.
Where people lose it: Candidates lose points by naming job titles without describing any verifiable workflow, responsibility or result.

Summary, eligibility guidance and screening preparation are prepared by Sidequest from the platform's posting. Check the source listing before applying.