AI Trainer (Speech-to-Speech Optimization) — Remote (June 2026 – Present)
Performed speech-to-speech optimization by evaluating and refining audio-to-audio LLM interactions for high-fidelity speech recognition and natural language generation. Conducted rigorous audio data labeling and transcription to improve model responsiveness and prosody in line with global quality benchmarks. Managed high-volume task pipelines to meet daily throughput targets while preserving expert-level accuracy in complex speech-based evaluation. • Labeled and transcribed audio data for model training improvements • Audited audio-to-audio interaction quality for recognition and generation fidelity • Maintained strict adherence to benchmark quality requirements • Tracked throughput and accuracy across daily evaluation batches