Yasir S.

Key Skills

Software

Don't disclose

Micro1

Top Subject Matter

Synthetic speech evaluation and RLHF-style preference testing for audio generation

Spoken language dataset creation for Malayalam conversational AI

Malayalam acoustic model training data for regional localization

Top Data Types

Audio

Video

Image

Top Task Types

Audio Recording

Transcription

Freelancer Overview

AI Audio Evaluation Consultant at Avyaan Management (2026). Brings 3+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Don't disclose and Micro1. Education includes Bachelor of Arts (Honours or Research), English and Foreign Languages University, Angadipuram (2021) and High School Diploma, Al Jamia Al Islamiya National Institute of Open Schooling, Santhapuram (2023). AI-training focus includes data types such as Audio and Video and labeling workflows including Evaluation, Rating, and Audio Recording.

Labeling Experience

Multimodal Data Annotator (English & Malayalam) at micro1 (2026)

Micro1VideoTranscription

Executed high-precision acoustic transcription and audio metadata logging for bilingual (English & Malayalam) video datasets. Documented multiple auditory elements including speech, background score, ambient environmental noise, and sound effects with precise time-stamps and quality metrics. Ensured richly annotated audio tracks to train computer vision and multimodal AI models using time-aligned auditory context. • High-precision acoustic transcription (English & Malayalam) • Time-stamped audio metadata logging for video segments • Labeled speech, background score, ambient noise, and sound effects • Added quality metrics to support multimodal training pipelines

2026

Language Consultant & Audio Contributor at Dip Audio (2026)

Don't discloseAudioAudio Recording

Recorded structured Malayalam conversational scripts to create clean native audio data for regional localization. Provided audio samples designed to capture regional phonetic nuances for training AI acoustic models. Supplied curated audio intended for use in acoustic model fine-tuning and localization workflows. • Recorded structured Malayalam conversational scripts • Captured regional phonetic nuances in native speech • Supplied clean audio for AI acoustic model training • Supported localization-focused dataset preparation

2026

Language Consultant & Audio Contributor at ChorusAudio (2026)

Don't discloseAudioAudio Recording

Contributed as an approved conversational speaker to record and provide high-quality spoken dataset audio. Supported dataset creation for Malayalam (India) language AI models to improve regional linguistic processing and conversational localization. Ensured the collected audio matched conversational dataset requirements for downstream training and tuning. • Recorded conversational utterances for Malayalam (India) • Provided clean spoken audio data for language model training • Supported regional linguistic processing advancements • Enabled conversational software localization

2026

AI Audio Evaluation Consultant at Avyaan Management (2026)

Don't discloseAudio

Performed human-in-the-loop binary preference testing on single-turn and multi-turn audio datasets to support AI training. Evaluated synthetic speech controllability and speaker consistency relative to target reference voices, validating audio quality independently of text. Logged and compared outcomes to determine whether generated audio outputs met reference-based criteria for training effectiveness. • Binary preference testing for audio pair comparisons • Controllability checks against target voice characteristics • Speaker consistency evaluation across multi-turn settings • Reference-voice based validation independent of textual content

2026

Education

Work History