Alex Moore

Key Skills

Software

Scale AI

Other

Snorkel AI

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Document

Image

Text

Top Label Types

Classification

Data Collection

Emotion Recognition

Entity Ner Classification

Evaluation Rating

Fine Tuning

Prompt Response Writing SFT

Question Answering

Red Teaming

Relationship

RLHF

Segmentation

Text Generation

Text Summarization

Transcription

Freelancer Overview

I am an experienced AI training expert and data annotation specialist with a strong background in training and evaluating large language models across a wide range of domains, including creative writing, history, biology (with a focus on microbiology, cell biology, and biophysics), and general content creation. My work involves crafting high-level prompts, ranking and rating AI-generated responses for factuality, safety, formatting, and style, as well as red-teaming and rubric creation to ensure high-quality outputs. I have contributed to projects for leading companies such as Snorkel AI, micro1, RemoExperts, Mercor, RWS & Alignerr, Outlier AI, and Appen, where I maintained outstanding quality standards and was recognized for exceptional performance. I am proficient in digital content development, CMS platforms, Python, SQL, and HTML5, and have applied my skills to tasks like transcription, speaker diarization, database management, and adapting to diverse tone and style requirements. My ability to manage multiple projects while maintaining accuracy and adaptability makes me well-suited for data labeling and AI training data roles.

Labeling Experience

Advanced STEM and Linguistics Projects

Snorkel AITextQuestion AnsweringText Generation

For these two projects, I help create frontier multi-modal datasets that measure model performance on multi-step reasoning and computation tasks, which can be used for SFT and RLFT across STEM disciplines. Additionally, I also create Humanity's Last Exam (HLE)-style questions designed to benchmark the reasoning abilities of state-of-the-art language models across multiple academic and professional domains. The quality measures are what you might typically expect for project and rubric pairs, such as validity of facts, correctness of any computations, applied principles, theories, etc., and more.

For these two projects, I help create frontier multi-modal datasets that measure model performance on multi-step reasoning and computation tasks, which can be used for SFT and RLFT across STEM disciplines. Additionally, I also create Humanity's Last Exam (HLE)-style questions designed to benchmark the reasoning abilities of state-of-the-art language models across multiple academic and professional domains. The quality measures are what you might typically expect for project and rubric pairs, such as validity of facts, correctness of any computations, applied principles, theories, etc., and more.

2025

Video Diarization

OtherAudioEntity Ner ClassificationEmotion Recognition

Annotate videos, identify the different speakers that are audible on the video by analyzing the voices, marking the correct time for the transcription, as well as noting the characteristics of the person talking, such as their vocal delivery, and perceived emotional tone.

Annotate videos, identify the different speakers that are audible on the video by analyzing the voices, marking the correct time for the transcription, as well as noting the characteristics of the person talking, such as their vocal delivery, and perceived emotional tone.

2025

Prompt and Response Writing Projects (Various)

Scale AITextQuestion AnsweringText Generation

Each of these projects revolved around prompts and rubric pair writing. The majority of these dealed with challenging state-of-the-art (SOTA) language models by asking questions based on provided texts or documents. I crafted questions that were complex enough for at least two of the models to consistently answer incorrectly. After receiving the models’ responses, I analyzed their outputs and provided a written critique or evaluation. Others included variations of writing a prompt that makes a model think, creating a rubric to describe that perfect response, and then writing a perfect response from the ground up. A few projects dealt with real user requests, too. One unique project was one where, in addition to writing a prompt that makes a model think and creating a rubric, I needed to ensure that the models searched the web for hard-to-find information and failed. Those models struggled with niche subjects & slightly more obscure texts, documents, and websites.

Each of these projects revolved around prompts and rubric pair writing. The majority of these dealed with challenging state-of-the-art (SOTA) language models by asking questions based on provided texts or documents. I crafted questions that were complex enough for at least two of the models to consistently answer incorrectly. After receiving the models’ responses, I analyzed their outputs and provided a written critique or evaluation. Others included variations of writing a prompt that makes a model think, creating a rubric to describe that perfect response, and then writing a perfect response from the ground up. A few projects dealt with real user requests, too. One unique project was one where, in addition to writing a prompt that makes a model think and creating a rubric, I needed to ensure that the models searched the web for hard-to-find information and failed. Those models struggled with niche subjects & slightly more obscure texts, documents, and websites.

2024 - 2025

Project Static Steerability

Scale AITextRelationshipText Summarization

We were tasked with evaluating and comparing two model responses within a conversation between a user and an AI character with a certain personality. We rated the responses against various categories and redirected conversations with NSFW content or flagged those tasks.

We were tasked with evaluating and comparing two model responses within a conversation between a user and an AI character with a certain personality. We rated the responses against various categories and redirected conversations with NSFW content or flagged those tasks.

2024 - 2025

WFE Conversation Rewrites (and Reviews)

Scale AITextClassificationFine Tuning

This project involved dealing with handling PII and NSFW content, marking instances of these two types of content,editing text to remove instances of them. I was made a reviewer on this project, so I also reviewed other contributors' work to make sure it upheld quality standards.

This project involved dealing with handling PII and NSFW content, marking instances of these two types of content,editing text to remove instances of them. I was made a reviewer on this project, so I also reviewed other contributors' work to make sure it upheld quality standards.

2024 - 2024

Education

Work History