OpenTrain partners with leading AI companies to build safer, smarter systems.
We’re contracting aviation-experienced professionals to transcribe and quality-check air traffic control (ATC) radio communications and produce structured training data for AI.
What you’ll do
- Transcribe tower/pilot radio calls verbatim from noisy, real-world audio.
- Tag entities such as call signs, runway numbers, altitudes, headings, readbacks/corrections; flag unclear or corrupted audio per policy.
- Apply strict formatting and consistency rules; meet weekly throughput targets.
- (QC track) Audit segments, correct issues, enforce guidelines, and provide concise feedback.
Minimum qualifications
- Real aviation background required: current/former ATC, pilot, flight instructor, dispatcher, or advanced ATC/aerospace student with radio-telephony phraseology training.
- Proven understanding of ICAO/FAA phraseology and standard readback patterns.
- Excellent English listening and attention to detail.
- Quiet workspace, Google Chrome, stable internet, and wired/high-quality headphones.
- Willing to complete a short evaluation and pass calibration/consistency checks.
Contract & schedule
- Type: Contract (independent contractor), remote.
- Hours: <20 hrs/week; minimum 10 hrs/week; flexible scheduling.
- Duration: 3–6 months (ongoing project).
- Openings: ~150 annotators (labelers and a smaller QC cohort).
Compensation
- Labelers: piece-rate USD $0.0125 per accepted audio second.
- Tasks receive a single QC pass; payment is on accepted output.
(Project volumes ramp from ~32 audio hours in week 1 toward ~1,000 audio hours/week overall.)
About OpenTrain
OpenTrain engages domain experts to create and evaluate high-quality datasets that teach advanced AI to understand complex, real-world signals—like ATC radio.
Location (eligible only): Bangladesh, Egypt, India, Pakistan, Philippines, Sri Lanka.
We are only contracting candidates currently residing in these countries.