Hire this AI Trainer
Sign in or create an account to invite AI Trainers to your job.
Data Annotation Tech
Document
Image
Text
Classification
Evaluation/Rating
Fine-tuning
Prompt + Response Writing (SFT)
Question AnsweringWith nearly a year of experience as a freelance data annotator for Data Annotation Tech, I’ve contributed to 50+ projects focused on evaluating and testing large language models (LLMs). My work involves identifying errors, fine-tuning models to mitigate bias, and ensuring AI outputs meet precise requirements like word counts and accurate data trends. I have significant experience with image-related projects, applying calculation and statistical analysis to verify data accuracy. My background in social sciences gives me a keen eye for detecting stereotyping and offensive content, particularly in politically and socially sensitive contexts. I’ve worked on comparing developing models with established ones like ChatGPT, and on projects related to safety, bias detection, and crafting ideal AI responses. In addition to technical expertise, I have strong analytical and writing skills. I’m adept at evaluating and labeling data, conducting statistical analysis, and reviewing the work of other annotators on the platform to ensure high-quality, consistent outputs.
Data Annotation TechDocument
Text SummarizationEvaluation/RatingEvaluation of AI models ability to summarise, extracts information and answer questions about documents and large bodies of text. This includes includes emails, PowerPoint presentations and business literature.
Data Annotation TechImage
Object DetectionEvaluation/RatingBengal is a large scale image evaluation project which tests AI models on their ability to analyse and accurately classify aspects of an image. This includes calculation and reasoning abilities in relation to images of graphs and statistics, as the model produced consistent errors in this area.
Data Annotation TechTextClassificationEvaluation/RatingThe AI models responses are tested with adversarial prompts, these prompts will test for bias and safety violations. Models will be evaluated based on whether they adhere to safety guidelines and how well they perform generally.
Level 2 UK Short Course, Understanding Coding
Bachelors in Sociology, Social Science
