Nurul Nadiah Binti Sahire

Audio Transcriber & Segmenter

Chemin AI

Dec 2025 - Current (8 months)

• Advanced Audio Segmentation: Performed high-precision audio slicing based on speech boundaries and logical sentence structures, strictly adhering to the 0.2-second pause rule and 8-second maximum segment length. • Verbatim Transcription: Transcribed Malay and English audio word-for-word with 95% accuracy, capturing every utterance including approved tone words (e.g., ah, um, lah, leh), stuttering, and repetitions without autocorrecting the speaker. • Complex Speaker Tagging: Accurately identified and labeled diverse speaker roles, including Telephony (Client/Agent) and Non-Telephony (Speaker A/B/C) roles, while managing metadata for gender and language. • Technical Annotation & Tagging: Applied specialized tags for non-speech sounds (

Data Validation Executive (Contract)

Newfire Global Partners

Jun 2024 - Dec 2025 (1 year 7 months)

• Proactively identify, research, and source accurate data on various clubs from online platforms, databases, and other reliable sources to meet weekly collection targets. • Ensured that all collected data is complete, consistent, and free of errors by performing thorough validation checks and cleaning processes. • Maintained data integrity by cross-verifying information from multiple sources and resolving discrepancies. • Coordinated with team members to align data collection efforts with project requirements or client specifications.

e

Freelance Data Annotator – Visual Response Assessment

e2f, Inc.

Mar 2024 - Jun 2024 (4 months)

• Evaluated Al-generated responses for harmfulness, suitability and temporal relevance based on provided guidelines. • Provided detailed justifications for evaluation choices to support data integrity and model improvement. • Worked independently to meet daily task quotas with high accuracy and attention to detail. • Participated in training and calibration sessions to align with evolving project standards.

P

AI Annotator (Contract)

PACTERA EDGE TECHNOLOGIES MALAYSIA SDN. BHD.

May 2022 - Sep 2022 (5 months)

• Determined and analysed input data by drawing boxes to highlight areas of interest of images and labelled them by using the Computer Vision Annotation Tool (CVAT) with a 97% accuracy rate. • Successfully completed annotate over 200 images per day, exceeding the daily targets by 20%. • Collaborated with the quality control team to improve annotation accuracy by 5%. • Reported the annotation glitches to the development team, which resulted in the improvement of a new feature and a decrease in annotation time by 15%

Nurul Nadiah Binti Sahire

About

Detail

Timeline

Résumé