David B.

Key Skills

Software

Other

Remotasks

Top Subject Matter

LLM/AI model training (RLHF) text evaluation and classification

Computer vision dataset annotation (segmentation and classification)

Legal Services & Contract Review

Top Data Types

Text

Image

Document

Top Task Types

RLHF

Segmentation

Object Detection

Data Collection

Freelancer Overview

Over the past six months, I have worked as a freelance Data Labeller and AI Trainer with Outlier AI and Atlas Capture, handling RLHF text annotation, intent and sentiment classification, bounding box and polygon image labelling, audio transcription, and multi-pass quality assurance, consistently hitting top accuracy scores across high-volume projects. This builds on over two years of actively learning about artificial intelligence, its systems, and how modern AI models are built and evaluated. What sets me apart is my attention to detail, strong QA discipline, and C2-level English proficiency (EFSET), which makes me especially effective on nuanced text annotation tasks. I work from a fully equipped dedicated home workspace with dual monitors, high-speed internet, and power backup, built for reliable, uninterrupted remote output.

Labeling Experience

Data Labeller (Freelance) | Atlas Capture

OtherImageSegmentation

Freelance data labeller for Atlas Capture, producing computer-vision annotations for model training datasets. Completed bounding box, polygon segmentation, landmark tagging, and image classification with attention to project-specific labeling criteria. Conducted multi-pass quality assurance and documented edge cases to improve consistency across labeling campaigns. • Labeled visual data using bounding boxes, polygons, and landmark tags • Applied image classification labels for vision dataset creation • Reviewed ambiguous cases and resolved issues during QA passes • Documented annotation decisions and edge-case resolutions for the team knowledge base

2026 - 2026

AI Trainer / Data Labeller (Freelance) | Outlier AI

OtherTextRLHF

Freelance AI trainer and data labeller for Outlier AI, annotating and evaluating AI-generated text outputs to support RLHF workflows. Labeled large volumes of conversational and semantic data, including intents, entities, sentiments, and conversation flow structure. Performed peer review and corrections to maintain dataset integrity and benchmark accuracy across iterative training cycles. • Rated response quality for accuracy, relevance, harmlessness, and overall quality • Followed changing annotation guidelines and style guides across concurrent projects • Conducted quality checks of peer annotations and applied correction notes • Collaborated asynchronously with QA leads and project managers to deliver on schedule

2025 - 2026

Education

Work History