What the work involves
You record scripted Swiss German audio — conversational exchanges, instructional lines, narrative passages — following a detailed spec covering microphone distance, room noise floor, file format, and naming conventions. Sessions are self-directed: you receive scripts, record takes at home, check them against the quality guidelines, and upload. Expect requests for multiple takes of the same line with different emotional colouring, emphasis, or pacing, and expect consistency checks across sessions weeks apart — the same voice, the same accent placement, the same energy level.
The headline term is stated plainly in the listing: your voice may be cloned for the client's CX AI agent. Mercor states the recordings are for that internal agent only and will not be sold, licensed, or reused elsewhere. Read the consent and usage terms carefully before applying; this is not a role to take if you have reservations about synthetic reproduction of your voice.
What the screen looks for
- Verified nativeness and location. Native Swiss German, currently resident in Switzerland. Expect probing on which dialect region you speak and how you'd handle scripts written in Standard German orthography.
- Provable professional history. Voice acting, dubbing, narration, podcasting, or broadcast credits you can point to.
- Your actual setup. Microphone model, interface, treatment, ambient noise conditions — answered specifically, not generically.
- Direction-following. Whether you can hit a stated emotional target on request and reproduce it later.
- Informed consent to cloning. Clear, unhesitant agreement to the stated terms.
Logistics
Fully remote and asynchronous, recorded from your own space. The listing specifies 5–10 hours per week across the project duration, with consistency across sessions mattering more than volume in any single week. Observed pay for this listing sits in the $50–150/hr range; actual rates depend on experience, screening outcome, and project scope, and are not guaranteed.