David Wayas

Key Skills

Software

Mercor

Internal/Proprietary Tooling

Telus

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Image

Text

Video

Top Label Types

Relationship

Object Detection

Data Collection

Prompt Response Writing SFT

Classification

Text Generation

Evaluation Rating

Freelancer Overview

I am an experienced data professional with a strong background in data labeling, annotation, and AI training data workflows. I have prepared, labeled, and enriched large-scale structured and text datasets, ensuring high accuracy and consistency for machine learning model development. My experience includes conducting quality audits, validation checks, and guideline-based reviews to improve training data reliability and reduce annotation errors. I am skilled in Python, SQL, Pandas, and have worked with cloud platforms like AWS, as well as data visualization tools such as Power BI and Tableau. I have collaborated closely with AI and analytics teams to refine labeling schemas and data structures, and have performed exploratory data analysis to identify bias and edge cases, ultimately enhancing model performance. My work spans domains such as e-commerce, sales analytics, and behavioral data, and I am adept at building automated data pipelines and dashboards that drive actionable insights.

Labeling Experience

Speech Data Annotation & Audio Quality Review Project

TelusAudioQuestion AnsweringEmotion Recognition

I contributed to AI training workflows involving speech and audio datasets used for voice recognition and conversational AI systems. My work focused on reviewing audio samples, validating transcription accuracy, tagging speech characteristics, and identifying issues such as background noise, unclear pronunciation, or misaligned labels. I performed consistency checks across annotated audio batches, flagged problematic samples, and ensured alignment between audio content and text transcripts. I also supported dataset cleaning and validation processes using Excel-based tracking to maintain annotation quality and guideline compliance. This project involved reviewing large batches of audio data and maintaining high accuracy standards to support reliable speech model training.

I contributed to AI training workflows involving speech and audio datasets used for voice recognition and conversational AI systems. My work focused on reviewing audio samples, validating transcription accuracy, tagging speech characteristics, and identifying issues such as background noise, unclear pronunciation, or misaligned labels. I performed consistency checks across annotated audio batches, flagged problematic samples, and ensured alignment between audio content and text transcripts. I also supported dataset cleaning and validation processes using Excel-based tracking to maintain annotation quality and guideline compliance. This project involved reviewing large batches of audio data and maintaining high accuracy standards to support reliable speech model training.

2024 - 2025

AI Training Data Annotation & Quality Review Project

MercorTextRelationshipObject Detection

I worked on large-scale AI training data preparation and annotation projects supporting machine learning and NLP model development. My responsibilities included labeling structured and text datasets according to strict guidelines, validating annotations for consistency, and performing quality audits to reduce model training errors. I reviewed edge cases, corrected misclassified samples, and collaborated with model development teams to refine labeling schemas and feedback loops. I also supported dataset cleaning and preprocessing using Excel and Python, improving annotation accuracy and downstream model performance. The project involved thousands of data records across multiple annotation cycles, with a strong focus on precision, consistency, and adherence to client-defined labeling standards.

I worked on large-scale AI training data preparation and annotation projects supporting machine learning and NLP model development. My responsibilities included labeling structured and text datasets according to strict guidelines, validating annotations for consistency, and performing quality audits to reduce model training errors. I reviewed edge cases, corrected misclassified samples, and collaborated with model development teams to refine labeling schemas and feedback loops. I also supported dataset cleaning and preprocessing using Excel and Python, improving annotation accuracy and downstream model performance. The project involved thousands of data records across multiple annotation cycles, with a strong focus on precision, consistency, and adherence to client-defined labeling standards.

2023 - 2024

Machine Learning Dataset Enrichment & Bias Detection Project

Internal Proprietary ToolingTextClassificationText Generation

I contributed to a machine learning dataset enrichment project focused on improving model accuracy and reducing bias in training data. I labeled and structured behavioural and text data, corrected inconsistencies, and flagged problematic samples affecting model predictions. My work included reviewing datasets for annotation gaps, identifying edge cases, and improving class balance through structured tagging. I also supported exploratory analysis to detect bias patterns and inconsistencies, ensuring the dataset aligned with training requirements. This work supported the development of predictive models and recommendation systems by ensuring the underlying training data was clean, consistent, and correctly labeled.

I contributed to a machine learning dataset enrichment project focused on improving model accuracy and reducing bias in training data. I labeled and structured behavioural and text data, corrected inconsistencies, and flagged problematic samples affecting model predictions. My work included reviewing datasets for annotation gaps, identifying edge cases, and improving class balance through structured tagging. I also supported exploratory analysis to detect bias patterns and inconsistencies, ensuring the dataset aligned with training requirements. This work supported the development of predictive models and recommendation systems by ensuring the underlying training data was clean, consistent, and correctly labeled.

2022 - 2023

Education

Work History