Caitlin Loehr

Key Skills

Software

Appen

Remotasks

Internal/Proprietary Tooling

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Top Task Types

Entity Ner Classification

Evaluation Rating

Prompt Response Writing SFT

RLHF

Text Summarization

Freelancer Overview

With exceptional skills in text analysis and writing, I have contributed to a variety of AI training tasks for prominent global clients, ranging from performing reinforcement learning through human feedback (RLHF); annotating training data; responding to prompts with original writing on diverse topics and in various styles; and reviewing tasks completed by other freelancers for quality control. My background in communication, behavioral science, and implementing technology projects globally equips me with a deep understanding of the factors that drive or dissuade people from using new technology, and I bring this human-centered perspective to my AI training work in order to ensure AI tools are user-friendly.

Labeling Experience

Generating Reference Summaries

AppenTextText Generation

Produced reference summaries to train AI models on generating summaries that align with human preferences.

Produced reference summaries to train AI models on generating summaries that align with human preferences.

2024 - 2024

Assessing Image Relevance

Scale AIImageEvaluation Rating

Evaluated responses that included images to determine relevance and usefulness of images.

Evaluated responses that included images to determine relevance and usefulness of images.

2024 - 2024

PII Identification

Scale AITextEntity Ner Classification

Identified Personally Identifiable Information (PII) in user prompts to support the AI model with recognizing this type of information and to protect user privacy by ensuring that identifying information would not be used for model training. Also reviewed PII classification done by other freelancers for quality control before completed tasks were sent to the client.

Identified Personally Identifiable Information (PII) in user prompts to support the AI model with recognizing this type of information and to protect user privacy by ensuring that identifying information would not be used for model training. Also reviewed PII classification done by other freelancers for quality control before completed tasks were sent to the client.

2024 - 2024

Response Evaluation

Scale AITextRLHFEvaluation Rating

Assessed multiple AI model outputs in response to the same prompt and justified which response was better according to various project criteria. Also reviewed evaluations and justifications of other freelancers for quality control before being sent to the clients.

Assessed multiple AI model outputs in response to the same prompt and justified which response was better according to various project criteria. Also reviewed evaluations and justifications of other freelancers for quality control before being sent to the clients.

2023 - 2024

Tagging Data for Training New NLP Model

Internal Proprietary ToolingTextClassificationData Collection

Tagged user prompts and scripted responses to frequently asked questions in order to begin the training process for a new NLP model that would be used to provide information and support to youth on sexual health topics.

Tagged user prompts and scripted responses to frequently asked questions in order to begin the training process for a new NLP model that would be used to provide information and support to youth on sexual health topics.

2013 - 2014

Education

Work History