Humphrey O.

Key Skills

Software

Sama

Top Subject Matter

Customer service chatbot retraining

conversational AI

safety and policy compliance

Top Data Types

Text

Document

Top Task Types

Entity (NER) Classification

Freelancer Overview

Data Annotation & AI Support Intern (Conversational prompt evaluation, labeling, and QA). Core strengths include Internal, Proprietary Tooling, and Sama. Education includes Bachelor of Science, Jomo Kenyatta University of Agriculture and Technology (JKUAT) (2023). AI-training focus includes data types such as Text and labeling workflows including Evaluation, Rating, and Entity (NER) Classification.

Labeling Experience

Data Annotation & AI Support Intern (Conversational prompt evaluation, labeling, and QA)

Text

Evaluated and annotated conversational prompts in English and Swahili to support customer service chatbot retraining. Performed quality assurance checks for grammatical accuracy, logical consistency, and policy compliance across training datasets. Conducted response and data error analysis to support model optimization and alignment improvements. • Reviewed and refined labeled prompt data for chatbot intent recognition • Assessed safety, toxicity, and policy adherence within training corpora • Collaborated to reduce biased and low-quality content via automated filtering • Produced feedback loops for ongoing AI optimization and evaluation

2024 - Present

Freelance AI Data Annotator (Semantic labeling and LLM response ranking)

SamaTextEntity (NER) Classification

Performed semantic annotation and categorization on large-scale text datasets for global AI projects. Maintained high quality accuracy across annotation assignments and contributed to building datasets for AI training and evaluation. Evaluated and ranked multi-turn LLM responses using truthfulness, helpfulness, coherence, and safety criteria, documenting edge cases to maintain annotation consistency. • Conducted semantic analysis and label assignment for dataset entries • Ranked multi-turn responses against truthfulness and safety criteria • Maintained ~98.5% average accuracy on annotation tasks • Authored guidance for ambiguous prompts and edge cases for consistency

2023 - 2023

Education

Work History