For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

LoginCreate account
Invite to Job
Jonathan Roos

Jonathan Roos

AI Data Evaluator - Technology & Internet

NETHERLANDS flag
Barendrecht, Netherlands
$17.00/hrEntry LevelInternal Proprietary Tooling

Key Skills

Software

Internal/Proprietary Tooling

Top Subject Matter

No subject matter listed

Top Data Types

VideoVideo
TextText

Top Label Types

Segmentation
Classification
Transcription
Evaluation Rating

Freelancer Overview

I am a detail-oriented AI data evaluator with hands-on experience in LLM output review, annotation, and quality assurance. My background in physics and data science helps me assess logical consistency, factual accuracy, and detect errors or hallucinations in AI-generated content. I have training in structured analytic techniques for systematic evaluation and bias reduction. I have previously annotated municipal council recordings following strict accuracy guidelines, maintaining consistent output quality under review. I have tested and graded prompt-response pairs across platforms including Amazon Bedrock and Claude, and am able to perform multilingual evaluation tasks in both Dutch (native) and English (fluent).

Entry LevelEnglishDutch

Labeling Experience

LLM Output Evaluation & Prompt Testing

Internal Proprietary ToolingTextEvaluation Rating
Evaluated LLM-generated outputs for logical consistency, factual accuracy, and hallucination detection across AI platforms including Amazon Bedrock and Claude. Tested and graded prompt-response pairs in business, workflow, logistics, and healthcare contexts. Reviewed AI-generated content for general digital assistant use cases. Identified errors, inconsistencies, and edge cases. Performed quality checks in both Dutch and English. Applied structured evaluation criteria to assess output quality systematically.

Evaluated LLM-generated outputs for logical consistency, factual accuracy, and hallucination detection across AI platforms including Amazon Bedrock and Claude. Tested and graded prompt-response pairs in business, workflow, logistics, and healthcare contexts. Reviewed AI-generated content for general digital assistant use cases. Identified errors, inconsistencies, and edge cases. Performed quality checks in both Dutch and English. Applied structured evaluation criteria to assess output quality systematically.

2023

Municipal Council Meeting Annotation & Indexing

Internal Proprietary ToolingVideoSegmentationClassification
Annotated and indexed municipal council meeting recordings for NotuBiz, a provider of council information systems for Dutch municipalities. Tasks included precise identification and tagging of speakers, agenda items, topics, and timestamps within audio/video recordings. Followed strict accuracy guidelines and structured criteria. Maintained consistent output quality across large volumes of content under quality review. Work required sustained focus and attention to detail over extended sessions.

Annotated and indexed municipal council meeting recordings for NotuBiz, a provider of council information systems for Dutch municipalities. Tasks included precise identification and tagging of speakers, agenda items, topics, and timestamps within audio/video recordings. Followed strict accuracy guidelines and structured criteria. Maintained consistent output quality across large volumes of content under quality review. Work required sustained focus and attention to detail over extended sessions.

2018 - 2018

Education

L

Leiden University

Bachelor of Science (ongoing), Physics (ongoing)

Bachelor of Science (ongoing)
2024 - 2025
M

MIT Professional Education

Professional Certificate, Applied Data Science

Professional Certificate
2023 - 2023

Work History

S

Self-employed

Freelance AI & Digital Projects

Rotterdam
2021 - Present
K

Koopman

Commercial & Logistics Specialist

Barendrecht
2019 - 2021