Dense instructions don’t scare me.
Current AI model-training work requires adapting to task-specific rules, quality requirements and changing evaluation criteria while working independently.
AI model training contractor, psychology student, C2-certified English communicator and systems-minded writer. I’m strongest where instructions are dense, edge cases matter, and the work needs both judgment and structure.
I’m early-career, not pretending otherwise. The advantage is that I already work like somebody who expects the brief to be complicated: I read closely, learn fast, document what matters, and don’t need to be handheld through every unfamiliar tool.
Current AI model-training work requires adapting to task-specific rules, quality requirements and changing evaluation criteria while working independently.
Cambridge C2 Proficiency at Grade A, long-form fiction, editing and research synthesis mean I can handle nuance, tone and precision instead of just producing grammatically acceptable text.
From canon databases to product concepts and game rules, I naturally turn messy information into structures people can navigate, maintain and use.
I like QA-style work because the interesting part is usually where the obvious rule stops being obvious. That makes me useful in evaluation, review and consistency-heavy workflows.
I’m especially suited to remote roles that combine language, structured reasoning, research, quality control or information-heavy creative work.
Rubric interpretation, output evaluation, annotation, quality review and careful instruction following.
See capability ↗02Source comparison, verification, analytical reading, concise synthesis and structured notes.
See capability ↗03Editing, proofreading, rewriting, tone consistency, documentation and long-form quality control.
See capability ↗04Information architecture, UX reasoning, structured data, concept development and documentation.
See capability ↗The portfolio is intentionally broad. The common thread is making complicated things coherent — whether the thing is an AI rubric, a fictional universe, a product concept or a research question.
A large-scale narrative system: interconnected countries, characters, politics, divine systems, timelines and continuity documentation.
Remote contract work under detailed task requirements, quality standards and confidentiality obligations.
A productivity concept translating an abstract behavioural idea into a structured feature system and user workflow.
I don’t need the task to match a job title I’ve already had. Give me the goal, constraints and examples; I’ll learn the system, ask useful questions when needed, and make the output cleaner than I found it.
Map the actual problem, assumptions and constraints before rushing into output.
Make decisions and rules explicit enough that someone else can follow them later.
Consistency breaks at the edges. I actively look for those breaks.
Complexity can stay backstage. The user or reader should get clarity.
If the job involves language, judgment, research, systems or quality, there’s a good chance I can become useful very quickly.