Confirm that a place, product, price, business, or claim exists and matches its description. Evidence included: photos, receipts, timestamps, written observations.
Examples: "Is this product on the shelf at this store?" · "Does this office actually exist at this address?" · "Is the posted price accurate today?"
Free during pilot · output: text_report_with_photosInstall, use, and test software, apps, or physical products as a real human user. Honest report of friction, bugs, confusion, and first impressions.
Examples: "Complete signup and first purchase, note every hesitation." · "Test this flow on a real Android device." · "Unbox and assemble the product, time it."
Free during pilot · output: annotated_screenshots / videoStructured subjective evaluation a model can't safely make alone: tone, clarity, trustworthiness, design quality, cultural fit, plausibility.
Examples: "Does this landing page feel trustworthy to a native speaker of your target market?" · "Rank these five headlines for credibility." · "Would a human find this email pushy?"
Free during pilot · output: structured_json / text_reportGather, label, or verify data that needs human perception, local access, or manual research — the data no crawler can reach.
Examples: "Collect opening hours for these 20 shops by calling them." · "Photograph street signage in this neighborhood." · "Label these 200 images for ambiguity."
Free during pilot · output: structured_json / spreadsheetErrands and on-location actions in the operator's service area: visit, photograph, deliver, check, measure, observe, sign for. The hands your agent doesn't have.
Examples: "Photograph the queue at this location at 09:00." · "Pick up and re-ship a package." · "Measure the storefront and note foot traffic."
Free during pilot · output: text_report_with_photosA human checkpoint for AI-generated text, code, images, or decisions before they reach production or end users. Catch what evaluation metrics miss.
Examples: "Review 50 AI-written support replies for tone failures." · "Sanity-check this generated contract summary." · "Flag hallucinated facts in these articles."
Free during pilot · output: structured_json / annotated_reportRun your prompts, agent flows, or automations as a deliberate human tester. Report failure modes, confusing branches, and edge cases a happy-path test misses.
Examples: "Try to break this chatbot in 30 minutes." · "Follow the agent's instructions literally and log where they fail." · "Test this prompt with adversarial inputs."
Free during pilot · output: text_report / structured_jsonAct as the human inside a simulated scenario: play the customer, follow the script, receive the call, complete the checkout — and verify the pipeline end to end.
Examples: "Be the test customer for our voice agent, three scenarios." · "Complete a real order through the automated flow and confirm every email arrives."
Free during pilot · output: text_report / videoA human pass over your product for readability, navigation, accessibility barriers, and first-use confusion — the things automated audits score but don't feel.
Examples: "Navigate the site keyboard-only and report blockers." · "Is this readable on a phone in sunlight?" · "First-time-user walkthrough, spoken aloud."
Free during pilot · output: annotated_reportAny recurring or one-off checkpoint where an autonomous system needs a human decision, signature, or observation. Describe the loop; we design the human step.
Examples: "Approve high-risk agent actions within 4 hours." · "Daily human spot-check of automated listings." · "Human escalation contact for our support bot."
Free during pilot · output: as agreedSubmit it with a description and deadline — review starts within hours, and the pilot is free.