
Amharic Speakers: Audio Transcription and Segmentation Project
Listing checked November 13, 2026 · details as published by Terac
Key details
- Pay
- Not published
- Location
- RemoteEthiopia, Eritrea, United States
- Job type
- freelance
- Openings
- 196
- Posted
- November 13, 2026
Overview
Terac runs a paid study that sharpens how well Amharic speech recognition models handle real recorded speech. Your day centers on a web platform where you listen to Amharic audio, then line up audio segmentation boundaries so each clip matches what the speaker actually says. Alongside that, you read the supplied transcripts and repair errors, misspellings, and misheard words. Steady contributions here feed straight into language tools that more people rely on. The project runs on an ongoing basis with a floor of 10 hours of work each week.
What You'll Do6
- Play through assigned Amharic audio clips and follow the speech as you go
- Check existing segment boundaries and shift them until they sit at the true start and end of each spoken phrase
- Read the provided transcripts against the audio and fix words that do not match
- Correct spelling and transcription slips so the written text mirrors what the speaker said
- Flag passages where the audio is unclear or the segmentation needs a second look
- Log at least 10 hours each week on the platform to keep tasks moving
Requirements6
- Fluent command of spoken Amharic
- Solid written Amharic, including accurate spelling
- Keen ear for subtle differences in recorded speech
- Comfort with browser-based annotation and text editing tools
- Willingness to commit a minimum of 10 hours per week
- Careful, repetitive work habits that hold up across long audio sessions
Preferred Qualifications3
- Background in transcription
- Background in translation
- Background in linguistics
Who Should Apply
Fluent Amharic speakers who catch small audio details and enjoy cleaning up text will settle in fast here. People with transcription, translation, or linguistics experience bring a useful head start. The role suits anyone who can protect 10 hours a week for ongoing annotation work.
Salary Insight
The listing does not publish a rate. Terac describes this as a paid study, so pay details come up during the application or onboarding conversation. Ask about the per-hour or per-task structure before you commit to the weekly 10-hour schedule.
Required skills
How applying works
1. Pre-screen on NearSkill
A quick application on NearSkill checks your fit against this role.
2. Finish on the employer’s site
We hand you to Terac’s official application page to complete the form.
3. Wait for a response
The employer’s hiring team reviews applications and gets back to you.
Similar jobs
Open roles that match this listing's skills and level.

TeracVerified
Posted 1d ago
Kinyarwanda Speakers: Audio Transcription for Speech Recognition
Pay not published
View job
AfterQueryVerified
Posted 2mo ago
Bilingual Audio Contributor - Indian Languages
$30/hr
View job
micro1Verified
Posted 19h ago
English Language Expert (United States)
$20–45/hr
View job
TuringVerified
Posted 25d ago
Music and Audio experts
$0.2K
View job
micro1Verified
Posted 13d ago
Music & Sound Annotation Expert
$30–45/hr
View job
micro1Verified
Posted 1mo ago
German Transcription Expert for Remote Contract Role
$20–30/hr
View job
