Håkon Bakke

Key Skills

Software

Scale AI

Labelbox

Top Subject Matter

Cybersecurity Domain Expertise

LLM Quality

AI Safety

Top Data Types

Text

Document

Top Task Types

Prompt + Response Writing (SFT)

Computer Programming/Coding

Evaluation/Rating

RLHF

Question Answering

Freelancer Overview

Cybersecurity professional and high-output AI trainer with hands-on experience across mathematics, Python/code evaluation, LLM assessment, and AI safety projects on Outlier (Scale AI). Over the course of 500+ completed tasks and 47 milestone rewards, I have written and refined prompts, solved problems independently to verify AI correctness, rated competing model outputs, and provided structured feedback on response quality — earning over $10,700 USD across active project periods. I bring genuine domain expertise to every evaluation rather than surface-level review. My background includes a Bachelor's in Cyber Security (Noroff University College) and an ongoing Master's in Cyber Security at Kristiania (GPA 4.5), covering threat intelligence, cryptography, AI for security, secure software development, and cloud security. This makes me particularly well-suited for evaluating AI-generated content in technical and security domains, identifying subtle errors that generalist reviewers would miss. I also hold a Microsoft Azure Fundamentals certification and have practical experience in SOC operations and physical security at a data centre.

Labeling Experience

AI Trainer & Domain Expert — Outlier (Scale AI)

Scale AITextRLHF

Worked as an AI trainer and domain expert on the Outlier (Scale AI) platform, specializing in the evaluation and improvement of AI-generated content. Assessed and rated the accuracy, quality, and safety of large language model (LLM) outputs, spanning domains such as cybersecurity, code review, mathematics, and general AI response evaluation. Provided structured feedback, comparative output ranking (RLHF), and technical quality reviews for AI-generated text and code. • Conducted detailed LLM response assessments for correctness, reasoning, and instruction relevance. • Evaluated code solutions, cybersecurity recommendations, and mathematical problem-solving. • Reviewed and rated safety and ethical alignment of AI outputs, identifying edge cases and failure modes. • Delivered over 500+ completed labeling and evaluation tasks using Outlier (Scale AI) as proprietary tooling.

Worked as an AI trainer and domain expert on the Outlier (Scale AI) platform, specializing in the evaluation and improvement of AI-generated content. Assessed and rated the accuracy, quality, and safety of large language model (LLM) outputs, spanning domains such as cybersecurity, code review, mathematics, and general AI response evaluation. Provided structured feedback, comparative output ranking (RLHF), and technical quality reviews for AI-generated text and code. • Conducted detailed LLM response assessments for correctness, reasoning, and instruction relevance. • Evaluated code solutions, cybersecurity recommendations, and mathematical problem-solving. • Reviewed and rated safety and ethical alignment of AI outputs, identifying edge cases and failure modes. • Delivered over 500+ completed labeling and evaluation tasks using Outlier (Scale AI) as proprietary tooling.

2025 - Present

Education

Work History