Murtadah Alamer

Key Skills

Software

Clickworker

CVAT

OneForma

SuperAnnotate

Supervisely

Toloka

Other

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Document

Image

Top Task Types

Audio Recording

Bounding Box

Classification

Data Collection

Polygon

Freelancer Overview

An AI Training Data Specialist with a background in Electrical Engineering, I have a proven track record of creating high-quality, diverse, and multilingual datasets that power AI. My expertise spans the full data lifecycle, from collection and multi-platform annotation to meticulous Quality Assurance. My unique value lies in four key areas: 🧠 Analytical & Mathematical Rigor: Leveraging a Bachelor's in Electrical Engineering from KFUPM and a Calculus Award, I apply strong analytical and logical reasoning to ensure high data integrity, accuracy, and consistency in every task. 🗣️ Native Arabic Linguist: Native fluency in Saudi Gulf dialect, strong comprehension of major Arabic dialects (Egyptian, Levantine, Hijazi, Najdi, Yamani), and university distinction in Modern Standard Arabic (MSA). I provide nuanced cultural and linguistic expertise for world-class Arabic language models. 🌐 Multimodal Data Annotation: Experience across data types, including Text, Image, Audio, Video, and 3D LiDAR point clouds. Familiar with platforms like AWS SageMaker, Azure ML, CVAT, Supervisely, UDT, and Roboflow, applying techniques from bounding boxes to complex polygon segmentation. 🔍 Data Quality Assurance: Achieved 96.6% task acceptance rate across 527 submissions on Hive Micro and 100% success in a QA review role. I excel at auditing data against complex rubrics to ensure accuracy and consistency.

Labeling Experience

Gulf Arabic Speech Transcriber at Sigma AI

OtherAudioText Generation

Transcribed Gulf Arabic and MSA audio in full verbatim for speech recognition models, accurately labeling speakers, timestamps, and overlapping dialogue. Scope included ongoing freelance work since Oct 2025, focusing on nuanced dialect comprehension to ensure high data accuracy for AI training.

Transcribed Gulf Arabic and MSA audio in full verbatim for speech recognition models, accurately labeling speakers, timestamps, and overlapping dialogue. Scope included ongoing freelance work since Oct 2025, focusing on nuanced dialect comprehension to ensure high data accuracy for AI training.

2025

Arabic Audio Data Specialist (Personal Project)

OtherAudioText Generation

Project Goal: To master the professional standards of audio transcription and QA by applying industry-leading guidelines to a variety of Arabic audio samples. Executed clean verbatim transcription of a 3-minute, Saudi Arabic interview, meticulously adhering to GoTranscript's style guide to achieve a self-assessed 98% accuracy score. Performed a critical QA role by auditing and correcting 10 AI-generated (HappyScribe) transcripts in both MSA and Lebanese Arabic, reformatting them to meet the rigorous GoTranscript quality standards for STT model training. Skills: Audio Transcription (Arabic) · Transcription QA · Dialect Comprehension (Saudi, MSA, Lebanese) · Guideline Adherence · Attention to Detail · Timestamping · Self-Assessment & Quality Control

Project Goal: To master the professional standards of audio transcription and QA by applying industry-leading guidelines to a variety of Arabic audio samples. Executed clean verbatim transcription of a 3-minute, Saudi Arabic interview, meticulously adhering to GoTranscript's style guide to achieve a self-assessed 98% accuracy score. Performed a critical QA role by auditing and correcting 10 AI-generated (HappyScribe) transcripts in both MSA and Lebanese Arabic, reformatting them to meet the rigorous GoTranscript quality standards for STT model training. Skills: Audio Transcription (Arabic) · Transcription QA · Dialect Comprehension (Saudi, MSA, Lebanese) · Guideline Adherence · Attention to Detail · Timestamping · Self-Assessment & Quality Control

2025 - 2025

AI Training Data Specialist (Project-Based Training)

CVATImageBounding BoxPolygon

Completed comprehensive project-based training in data annotation for computer vision and NLP. Gained hands-on experience in full lifecycle: setup, annotation, QA, using tools like AWS SageMaker, Azure ML, CVAT, Roboflow. Key accomplishments: - Multi-platform proficiency: Adapted to UIs/workflows for end-to-end projects. - CV annotation: Labeled 12-image dataset (80+ cars via bounding boxes, polygons); classified 100+ boot/sneaker images; 45-image single-label project. - NLP/audio: NER on text (15+ entities like ORG/PERSON), sentiment analysis; transcribed/diarized multi-language audio. - Model-assisted: Used YOLOv7 in CVAT, Label Assist in Roboflow, SAM for efficient semi-automated labeling. - Exported datasets in COCO (JSON), Pascal VOC (XML), YOLO (TXT). Skills: Cloud platforms (AWS SageMaker, Azure ML) · Tools (CVAT, Roboflow, UDT) · CV (detection, segmentation) · NLP (NER, sentiment) · Model-assisted labeling · Formats (COCO, YOLO).

Completed comprehensive project-based training in data annotation for computer vision and NLP. Gained hands-on experience in full lifecycle: setup, annotation, QA, using tools like AWS SageMaker, Azure ML, CVAT, Roboflow. Key accomplishments: - Multi-platform proficiency: Adapted to UIs/workflows for end-to-end projects. - CV annotation: Labeled 12-image dataset (80+ cars via bounding boxes, polygons); classified 100+ boot/sneaker images; 45-image single-label project. - NLP/audio: NER on text (15+ entities like ORG/PERSON), sentiment analysis; transcribed/diarized multi-language audio. - Model-assisted: Used YOLOv7 in CVAT, Label Assist in Roboflow, SAM for efficient semi-automated labeling. - Exported datasets in COCO (JSON), Pascal VOC (XML), YOLO (TXT). Skills: Cloud platforms (AWS SageMaker, Azure ML) · Tools (CVAT, Roboflow, UDT) · CV (detection, segmentation) · NLP (NER, sentiment) · Model-assisted labeling · Formats (COCO, YOLO).

2025 - 2025

Video Timestamping Specialist at Hive Micro

Internal Proprietary ToolingVideoClassification

Created diverse training data for computer vision models at Hive Micro. Tasks included video timestamping of 32 segments (1.5 mins/task) with precise, frame-level intro/outro timestamps and multi-level content classification (e.g., Commercial, TV Promo) based on hierarchical rubrics.

Created diverse training data for computer vision models at Hive Micro. Tasks included video timestamping of 32 segments (1.5 mins/task) with precise, frame-level intro/outro timestamps and multi-level content classification (e.g., Commercial, TV Promo) based on hierarchical rubrics.

2025 - 2025

Video Data Collector for AI Models

ClickworkerVideoObject DetectionData Collection

Executed data collection for AI/ML models at Clickworker, including short videos for facial recognition and gesture analysis per precise specs. Adhered to strict guidelines for quality, privacy, and security to produce high-quality training data.

Executed data collection for AI/ML models at Clickworker, including short videos for facial recognition and gesture analysis per precise specs. Adhered to strict guidelines for quality, privacy, and security to produce high-quality training data.

2025 - 2025

Show more

Education

Work History