Harris K.

Key Skills

Software

Appen

Other

Top Subject Matter

LLM evaluation and supervised fine-tuning data annotation

Search ranking evaluation and relevance labeling

Top Data Types

Image

Text

Video

Top Task Types

Bounding Box

Classification

Evaluation/Rating

Text Generation

Freelancer Overview

Data Annotation Specialist — Outlier AI (LLM data work, evaluation, and multimodal SFT support). Brings 9+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Internal, Proprietary Tooling, and Appen. Education includes Bachelor of Engineering (Honours), Loughborough University (2019) and Engineering (Initial Year), University of Hertfordshire (2016). AI-training focus includes data types such as Text and Image and labeling workflows including Evaluation, Rating, and Classification.

Labeling Experience

Data Annotation Specialist - Outlier AI

AppenImage

Produced high-quality annotated datasets used to train and evaluate large language models (LLMs). Evaluated AI outputs for factual accuracy, reasoning quality, and instruction adherence. Identified model failure modes and provided structured feedback to improve model performance. • Evaluated model reasoning, factual correctness, and instruction-following behaviour across diverse prompts. • Assisted with multimodal supervised fine-tuning tasks focused on image-to-text alignment. • Applied prompt engineering and annotation taxonomies aligned to model objectives. • Maintained strict quality standards in fast-paced AI training pipelines.

2026 - Present

Data Annotation Specialist — Outlier AI (LLM data work, evaluation, and multimodal SFT support)

Text

Produced high-quality annotated datasets to train and evaluate large language models (LLMs). Evaluated model outputs for factual accuracy, reasoning quality, and instruction-following behavior across diverse prompts. Identified model failure modes and delivered structured feedback to improve model performance. • Assessed factual correctness and reasoning quality• Graded prompt adherence and instruction following• Conducted failure-mode identification with actionable feedback• Supported multimodal supervised fine-tuning (image-to-text alignment) using annotation taxonomies

2026 - Present

AI data specialist - Annotation + More

VideoObject Detection

Produced high-quality annotated datasets used to train and evaluate large language models (LLMs). Evaluated AI outputs for factual accuracy, reasoning quality, and instruction adherence Identified model failure modes and provided structured feedback to improve model performance Evaluated model reasoning, factual correctness, and instruction-following behaviour across diverse prompts Applied prompt engineering and annotation taxonomies to align training data with model objectives Maintained strict quality standards while working independently in fast-paced AI training pipelines

2026 - Present

SFT Prompt Writing

ImagePrompt + Response Writing (SFT)

Worked on supervised fine-tuning (SFT) and RLHF pipelines for large language models, producing high-quality annotated datasets used to train and evaluate frontier AI systems. Tasks included evaluating LLM outputs for factual accuracy, reasoning quality, coherence, and instruction-following behaviour across diverse prompt types. Identified and documented model failure modes including hallucinations, refusals, and reasoning errors. Applied prompt engineering and annotation taxonomies to align training data with model objectives. Maintained strict quality standards while working independently at pace within fast-moving AI training pipelines.

2026 - Present

Search Evaluation Consultant - Appen

AppenImage

Evaluated search engine results and advertisements for relevance, quality, and user intent. Annotated datasets used to train and evaluate machine learning search ranking models. Performed query intent analysis and provided structured feedback to improve ranking algorithms. • Assessed content relevance and supported human-in-the-loop machine learning workflows. • Labeled and reviewed results according to labeling guidelines. • Contributed to improving ranking algorithms through feedback and dataset quality work. • Worked within large-scale evaluation pipelines under quality standards.

2018 - 2020

Show more

Education

Work History