Jessica Brown-Flemons

Key Skills

Software

Scale AI

Top Subject Matter

Machine Learning/AI Model Training

Natural Language Processing (NLP)/LLM Evaluation

Legal Services & Contract Review

Top Data Types

Image

Text

Document

Top Task Types

RLHF

Bounding Box

Object Detection

Freelancer Overview

During my time on the Aether project at Outlier, I specialized in high-level data annotation and model evaluation, focusing on the refinement of AI responses through Reinforcement Learning from Human Feedback. My work centered on judging the accuracy and safety of generated content, particularly within visual and text-based prompts. I was responsible for ensuring that the model followed complex instructions while maintaining a high standard for quality, which required a sharp eye for detail and a strong understanding of content guidelines. This person also contributed to the early stages of Google's Gemini model development. This role included human evaluation to improve the model's reasoning and multimodal capabilities. Tasks included testing the AI's response consistency and assessing performance across varied datasets, with a specific focus on image and video judging and prompt engineering. This experience helped develop proficiency in complex evaluation platforms, delivering quick and accurate results.

Labeling Experience

AI Data Specialist (Contractor/Freelance), Outlier.ai (via Scale AI Platform)

Scale AIImageRLHF

As an AI Data Specialist, I contributed to developing AI models through image and video evaluation, comparison, and annotation. I provided high-quality human feedback using Reinforcement Learning from Human Feedback (RLHF) to improve model accuracy and safety. My work included labeling features in visual content and ranking AI-generated responses for clarity and correctness. • Performed detailed annotation of video clips and screenshots to support visual model training. • Applied critical thinking and communication skills to report on edge cases and data inconsistencies. • Regularly evaluated the performance and naturalness of AI-generated outputs. • Managed deadlines and maintained accuracy throughout project durations.

As an AI Data Specialist, I contributed to developing AI models through image and video evaluation, comparison, and annotation. I provided high-quality human feedback using Reinforcement Learning from Human Feedback (RLHF) to improve model accuracy and safety. My work included labeling features in visual content and ranking AI-generated responses for clarity and correctness. • Performed detailed annotation of video clips and screenshots to support visual model training. • Applied critical thinking and communication skills to report on edge cases and data inconsistencies. • Regularly evaluated the performance and naturalness of AI-generated outputs. • Managed deadlines and maintained accuracy throughout project durations.

2025 - Present

AI Model Training & Feedback Program - Google Gemini Platform

Text

I participated in a feedback program to evaluate and enhance the performance of a large language model. Tasks included providing comprehensive feedback on AI-generated content to ensure accuracy, relevance, and safety. Analytical and troubleshooting skills were applied to assess outputs and support natural language processing (NLP) improvements. • Ranked diverse LLM outputs against strict project guidelines. • Edited and reviewed AI-generated text for correctness and naturalness. • Assisted in troubleshooting model errors and reporting issues. • Contributed feedback to inform ongoing improvements of AI systems.

I participated in a feedback program to evaluate and enhance the performance of a large language model. Tasks included providing comprehensive feedback on AI-generated content to ensure accuracy, relevance, and safety. Analytical and troubleshooting skills were applied to assess outputs and support natural language processing (NLP) improvements. • Ranked diverse LLM outputs against strict project guidelines. • Edited and reviewed AI-generated text for correctness and naturalness. • Assisted in troubleshooting model errors and reporting issues. • Contributed feedback to inform ongoing improvements of AI systems.

2025 - 2025

Education

Work History