Join Project Koala today.

At OneForma,​ we are inviting qualified contributors​ to participate​ in our new remote annotation project.

Purpose:

Project Koala supports the training and evaluation​ of advanced speech recognition systems. Participants will contribute​ by accurately identifying and labeling conversational events​ in recorded audio interactions between users and voice assistants.

Location:

Remote. Open​ to native-level speakers of:

Chinese (Hong Kong), Chinese (Taiwan), Chinese Simplified (China), Dutch (Belgium), Malay (Malaysia), Norwegian-Bokmal (Norway).

Please note: This project​ is not available​ in Nigeria.

Responsibilities:

Participants will listen​ to recorded conversations and apply precise, time-stamped labels to identify speakers and conversational events according​ to detailed project guidelines.

All selected contributors must:

 • Pass the OneForma Certification process

 • Complete KYC verification

 • Successfully pass required practice tasks within the client’s annotation tool before beginning production

Duration and Commitment:

Ongoing project. Participants are required to commit a minimum​ of​ 20 hours per week, subject​ to task availability and performance quality.

Compensation:

Payment​ is provided per approved audio hour​ or task. Rates will​ be shared with qualified candidates during the onboarding process.

A member​ of our team will contact qualified candidates with the next steps.

Main Requirements:

 • Native fluency​ in one​ of the target languages 

     - Dutch (Belgium)

     - Norwegian Bokmål (Norway)

     - Malay (Malaysia)

     - Chinese Simplified (China)

     - Chinese (Hong Kong)

     - Chinese (Taiwan)

 • Strong listening comprehension and ability to distinguish multiple speakers

 • Excellent attention​ to detail and ability to follow structured linguistic and technical guidelines

 • Reliable computer, stable internet connection, microphone, and headset

 • Comfortable using web-based annotation tools

 • Previous experience​ in annotation, transcription, linguistics,​ or data labeling​ is​ a plus, but not mandatory.