Tun Eaindar May

Burmese audio data professional with current hands-on experience reviewing and correcting ASR-generated transcripts for Perle AI. Experienced in audio segmentation, timestamp adjustment, speaker alignment, transcription correction, manual transcription, and identifying difficult audio conditions including background noise, overlapping speakers, dialects, accents, code-switching, and audio interruptions. Supported by 4+ years of professional quality assurance and computer vision data annotation experience, with a strong track record of following detailed guidelines, maintaining consistency, and handling high-volume quality-focused workflows.

Burmese Freelance Transcription Reviewer at Perle AI (2026-08 – Present)

• Review Burmese audio recordings and compare ASR-generated transcripts against the original audio to identify transcription and timing errors.
• Correct inaccurate ASR output and manually transcribe sections when the generated transcript is incomplete or significantly inaccurate.
• Create new transcript segments and adjust segment timestamps to ensure each speaker's speech is accurately aligned with the corresponding portion of the audio.
• Review speaker segmentation and ensure transcript segments correspond correctly to changes in speaker speech.
• Correct missing words, extra words, incorrect words, spelling, formatting, and other transcription inconsistencies.
• Handle challenging audio containing background noise, unclear speech, overlapping speakers, accents, regional dialects, Burmese-English code-switching, music or television in the background, and audio interruptions or dropouts.
• Tag unclear speech as [inaudible] when the spoken content cannot be reliably determined rather than guessing.
• Identify foreign-language speech and Burmese dialects and report them through the project communication channel with the relevant task ID, audio duration, and segment number.
• Regularly perform segment creation, splitting, merging, and timestamp adjustments when the ASR-generated segmentation does not accurately represent the audio.
• Follow detailed transcription and annotation guidelines consistently across project tasks.
• Use Lingo as the primary transcription review and annotation platform.
• Completed approximately 22 audio hours of Burmese transcription review to date.
• Maintain a minimum level of activity required by the project while working independently in a remote environment.

Quality Assurance Technician at Infinity Global Solutions (2017-01 – 2021-05)

• Reviewed and validated approximately 300 product image data items per day for international computer vision projects.
• Performed high-volume data annotation and quality assurance according to detailed project-specific guidelines.
• Identified incorrect labels, inconsistencies, annotation errors, and other quality issues.
• Corrected or flagged data that did not meet required quality standards.
• Applied annotation guidelines consistently across large volumes of visual data.
• Carefully compared source images with annotations to verify accuracy and completeness.
• Maintained productivity and accuracy while performing repetitive, detail-sensitive quality tasks.
• Adapted to changing project instructions and requirements while maintaining consistent quality.
• Worked remotely from home during the final two years of the position.

Bachelor of Science in Mathematics – Mandalay University of Distance Education