Nqobile S.

Key Skills

Software

Other

Top Subject Matter

Audio speech evaluation and transcription for AI model training

Multilingual response evaluation and language detection labeling

Safety review

Top Data Types

Audio

Text

Video

Document

Top Task Types

Transcription

Fine-tuning

Freelancer Overview

Audio Reviewer at VerityLab.AI (Audio model evaluation, transcription, and QA). Brings 15+ years of professional experience across legal operations, contract review, compliance, and structured analysis. Core strengths include Don't disclose, Other, and Internal. Education includes Bachelor’s Degree, Unisa (2016) and Short Course, Unisa (2015). AI-training focus includes data types such as Audio and Text and labeling workflows including Transcription, Evaluation, and Rating.

Labeling Experience

Audio Reviewer at VerityLab.AI (Audio model evaluation, transcription, and QA)

Don't discloseAudioTranscription

Reviewed and transcribed audio/ASR content while rating how closely model outputs match human speech. Performed evaluations on both synthetic audio model outputs and real human speech datasets to support better downstream understanding. Conducted quality assurance on annotator outputs to improve accuracy and consistency of training data. • Rated audio model outputs for closeness to human speech • Transcribed ASR data for researcher use in audio model training • Evaluated human speech samples for user understanding • Performed QA checks on annotator data for accuracy

2025 - Present

Advanced AI Data Trainer (Contractor) at Invisible Technologies

TextFine-tuning

Improved AI model accuracy and safety by reviewing and flagging harmful content using custom annotation platforms. Rated and ranked AI responses according to English language criteria and client guidelines to support dataset quality improvements. Verified that generated content was appropriate and useful for user prompts while streamlining high-volume evaluation workflows. • Reviewed content and flagged harmful responses using annotation platforms • Rated and ranked AI responses based on English criteria and guidelines • Ensured responses matched user prompts and were useful • Used browser-based tools to support high-volume remote annotations

2024 - 2025

Multilingual Evaluator (Contractor) at RWS Moravia

OtherText

Analyzed and classified multilingual responses by entering language detection results into centralized tracking systems. Identified additional languages present in responses and documented them to improve language recognition and response quality. Ensured consistent language tagging and evaluation accuracy using remote evaluation tools and cloud databases. • Classified multilingual responses and recorded language detection results • Identified additional languages and documented findings for training • Maintained consistency of language tagging and response evaluation • Used remote tools and cloud databases for evaluation tracking

2024 - 2025

Education

Work History