Jastrow Ogbu

Key Skills

Software

Appen

Clickworker

Google Cloud Vertex AI

Toloka

Telus

Other

Internal/Proprietary Tooling

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Video

Top Task Types

Bounding Box

Classification

Evaluation/Rating

Question Answering

RLHF

Freelancer Overview

I am an expert AI Data Trainer and Evaluator with over 4 years of hands-on experience refining Large Language Models (LLMs) and Computer Vision systems. My background combines specialized STEM knowledge in Physics and Engineering with extensive proficiency in computer programming (Python, HTML, CSS). I excel in high-complexity tasks such as RLHF (Reinforcement Learning from Human Feedback), where I evaluate model responses for accuracy, safety, and coherence, particularly in technical and scientific domains. I also have deep experience in Red Teaming and misinformation detection, ensuring AI outputs meet rigorous safety standards. Beyond text-based training, I am highly skilled in multi-modal data annotation. I have successfully completed large-scale projects involving bounding box and polygon annotation, as well as precise video tracking and audio transcription. My ability to handle technical code reviews and solve advanced mathematical problems sets me apart from general annotators. I am detail-oriented, adaptable to new guidelines, and committed to delivering high-quality training data that enhances AI performance across diverse industries.

Labeling Experience

Multi-Modal Video Q&A Evaluation & Hallucination Correction

AppenVideoQuestion AnsweringRLHF

Executed high-precision evaluation for a Multi-Modal Large Language Model (MLLM) focused on video understanding in daily life scenarios. The workflow involved 'Video Grounding,' where I analyzed video clips to verify the accuracy of AI-generated Question-and-Answer (Q&A) pairs.

Executed high-precision evaluation for a Multi-Modal Large Language Model (MLLM) focused on video understanding in daily life scenarios. The workflow involved 'Video Grounding,' where I analyzed video clips to verify the accuracy of AI-generated Question-and-Answer (Q&A) pairs.

2025

AI Safety & Red Teaming Specialist (Misinformation/Harmlessness)

AppenTextClassificationQuestion Answering

Operated as a Safety Evaluator and Red Teamer for Project Uolo, a critical initiative focused on detecting and mitigating misinformation, hate speech, and harmful content in Large Language Models.

Operated as a Safety Evaluator and Red Teamer for Project Uolo, a critical initiative focused on detecting and mitigating misinformation, hate speech, and harmful content in Large Language Models.

2022

Social Media Visual Search & Entity Tagging

TelusImageBounding BoxPolygon

Contributed to a large-scale visual search project for a major social media platform (Instagram integration). The objective was to train computer vision models to accurately identify, segment, and classify distinct entities within complex, real-world images.

Contributed to a large-scale visual search project for a major social media platform (Instagram integration). The objective was to train computer vision models to accurately identify, segment, and classify distinct entities within complex, real-world images.

2025 - 2025

Animal Data Collection & Health Classification (Computer Vision)

AppenImageClassificationEvaluation Rating

Executed a comprehensive data collection and annotation project aimed at training Computer Vision models for veterinary health assessment. The scope involved sourcing and analyzing a diverse dataset of canine imagery to automate Body Condition Scoring (BCS).

Executed a comprehensive data collection and annotation project aimed at training Computer Vision models for veterinary health assessment. The scope involved sourcing and analyzing a diverse dataset of canine imagery to automate Body Condition Scoring (BCS).

2025 - 2025

Advanced Mathematics Subject Matter Expert (LLM Training)

TelusTextQuestion AnsweringText Generation

Served as a Mathematics Subject Matter Expert (SME) to advance the reasoning capabilities of Large Language Models. The role focused on designing high-level problem sets that test deep reasoning and creating authoritative 'Golden Solutions' for model training.

Served as a Mathematics Subject Matter Expert (SME) to advance the reasoning capabilities of Large Language Models. The role focused on designing high-level problem sets that test deep reasoning and creating authoritative 'Golden Solutions' for model training.

2025 - 2025

Show more

Education

Work History