Stephen Foster

Key Skills

Software

Data Annotation Tech

Remotasks

Scale AI

Appen

Labelbox

Mercor

Top Subject Matter

Computer Vision and Visual Data Annotation

Image Generation and Prompt Design

Creative Applications of Artificial Intelligence in Art and Design

Top Data Types

Text

Top Label Types

Prompt Response Writing SFT

Freelancer Overview

I am an experienced Data Annotation Specialist with a strong background in visual storytelling, photography, and research. My work involves training AI models through precise data labeling, annotation, and optimization, with hands-on experience in chatbot interaction, content writing and editing, and coding tasks. I am adept at using advanced tools and platforms to enhance machine learning models, particularly in domains like computer vision and natural language processing. My creative and technical expertise, developed through years of working with diverse clients in tech, academia, and the arts, allows me to deliver high-quality, detail-oriented training data for AI applications. I thrive in multidisciplinary environments and am passionate about contributing to innovative AI solutions.

Labeling Experience

Professional Photography SFT & Visual Data Generation

MercorImageClassificationData Collection

Generated high-resolution, original photographic datasets specifically designed for model training. This role required capturing original imagery that aligned with complex creative briefs, which involved managing professional lighting, composition, and technical framing to ensure consistency across experimental visual formats. I also authored high-fidelity Supervised Fine-Tuning data, including technical descriptions of lighting physics, depth of field, and stylistic context to ground the model in professional photographic standards. Additionally, I managed the end-to-end data delivery process, including file tagging and structured categorization of visual content for research-driven AI applications.

Generated high-resolution, original photographic datasets specifically designed for model training. This role required capturing original imagery that aligned with complex creative briefs, which involved managing professional lighting, composition, and technical framing to ensure consistency across experimental visual formats. I also authored high-fidelity Supervised Fine-Tuning data, including technical descriptions of lighting physics, depth of field, and stylistic context to ground the model in professional photographic standards. Additionally, I managed the end-to-end data delivery process, including file tagging and structured categorization of visual content for research-driven AI applications.

2025 - 2025

Multi-Modal Red Teaming

Scale AIImageEvaluation RatingRed Teaming

Challenged language models by sourcing complex images and designing adversarial prompts to expose model limitations. Analyzed and critiqued model outputs where multiple models failed. Additionally, wrote ideal prompt-response pairs based on varied use cases to simulate high-quality chatbot behavior for Supervised Fine-Tuning.

Challenged language models by sourcing complex images and designing adversarial prompts to expose model limitations. Analyzed and critiqued model outputs where multiple models failed. Additionally, wrote ideal prompt-response pairs based on varied use cases to simulate high-quality chatbot behavior for Supervised Fine-Tuning.

2025 - 2025

Supervised Fine-Tuning

LabelboxTextPrompt Response Writing SFT

Executed Supervised Fine-Tuning tasks, specifically focusing on the "SFT Rework" pipeline, to improve language model reasoning and formatting. Crafted high-quality prompt and response pairs and refined existing datasets to ensure strict alignment with human preference guidelines.

Executed Supervised Fine-Tuning tasks, specifically focusing on the "SFT Rework" pipeline, to improve language model reasoning and formatting. Crafted high-quality prompt and response pairs and refined existing datasets to ensure strict alignment with human preference guidelines.

2024 - 2025

Contrastive Video Labeling

AppenVideoClassificationEvaluation Rating

Processed and analyzed short video clips via the SRT interface to perform AutoCapture contrastive labeling. Navigated high-volume production queues and strict time-per-unit limits to provide accurate video comprehension data for AI training models.

Processed and analyzed short video clips via the SRT interface to perform AutoCapture contrastive labeling. Navigated high-volume production queues and strict time-per-unit limits to provide accurate video comprehension data for AI training models.

2024 - 2025

Audio Persona Steerability Testing

Scale AIAudioEvaluation RatingPrompt Response Writing SFT

Conducted interactive, multi-turn conversations with an audio-based chatbot to test persona adherence and system prompt steerability. Transcribed audio interactions, developed custom grading rubrics for each turn, and evaluated model performance to ensure the AI accurately maintained its assigned character.

Conducted interactive, multi-turn conversations with an audio-based chatbot to test persona adherence and system prompt steerability. Transcribed audio interactions, developed custom grading rubrics for each turn, and evaluated model performance to ensure the AI accurately maintained its assigned character.

2024 - 2025

Show more

Education

Work History