For employers

Hire this AI Trainer

Sign in or create an account to invite AI Trainers to your job.

Invite to Job
Nancy Kinyua

Nancy Kinyua

AI Data Annotator -Text Classification and Rating

Kenya flagNairobi, Kenya
$30.00/hrExpertAppenOther

Key Skills

Software

AppenAppen
Other

Top Subject Matter

Customer Service -Chatbots &Support
E-commerce-product Descriptions & Reviews
Technology-AI & Machine learning

Top Data Types

TextText
DocumentDocument

Top Task Types

Evaluation Rating
Transcription
Classification

Freelancer Overview

Detail-oriented AI Data Annotator with strong skills in prompt analysis, response evaluation, and data labelling for machine learning systems. Experienced in reviewing AI-generated outputs to assess accuracy, reasoning quality, and instruction adherence. Skilled at following complex guidelines and maintaining high consistency when working with training datasets. Known for strong English comprehension, analytical thinking, and delivering reliable, high-quality work in remote AI training environments.

ExpertEnglishSwahili

Labeling Experience

Ranking Chatbot Responses for Helpfulness

TextEvaluation Rating
Compared two AI-generated answers to the same user question and selected which one was more helpful, accurate, and harmless.Labeled 500 prompts across topics like general knowledge, writing help, and basic math. Followed a simple rubric: helpful >vague, accurate> made-up info, safe>toxic.Weekly feedback sessions improved consistency.

Compared two AI-generated answers to the same user question and selected which one was more helpful, accurate, and harmless.Labeled 500 prompts across topics like general knowledge, writing help, and basic math. Followed a simple rubric: helpful >vague, accurate> made-up info, safe>toxic.Weekly feedback sessions improved consistency.

2024 - Present

Email Classification for Spam Detection

TextClassification
Labeled 2,000 emails as either "spam" or "not spam" to help train a simple email filter.Each email was reviewed and tagged using a basic labelling tool.Worked with a small team of 3 people. A second person double checked 10% of the emails to ensure accuracy.

Labeled 2,000 emails as either "spam" or "not spam" to help train a simple email filter.Each email was reviewed and tagged using a basic labelling tool.Worked with a small team of 3 people. A second person double checked 10% of the emails to ensure accuracy.

2024 - 2026

Education

N

NIBS Technical College

NIBS Technical College (no degree), Communications

NIBS Technical College (no degree)
2023 - 2024

Work History

C

CrowdGen by Appen

Speech Data and Transcription Contributor (Freelance)

Nairobi
2025 - Present