Mananki

Key Skills

Software

Mercor

Top Subject Matter

Natural Language Processing

Large Language Models

Multilingual AI

Top Data Types

Text

Document

Top Task Types

RLHF

Question Answering

Object Detection

Entity Ner Classification

Freelancer Overview

AI Data Trainer. Brings 1+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Mercor. Education includes Bachelor of Business Information Technology, Mount Kenya University (2021) and Kenya Certificate of Secondary Education, Nguviu Boys High School (2020). AI-training focus includes data types such as Text and labeling workflows including Evaluation and Rating.

Labeling Experience

AI Data Trainer

MercorText

As an AI Data Trainer at Mercor AI, I evaluated and improved large language model outputs to ensure logic, factual accuracy, and helpfulness. I conducted labeling and quality assurance through prompt engineering, multilingual dialogue annotation, and RLHF methodologies. My work enhanced model performance by identifying errors, biases, and supporting expansion to low-resource East African languages. • Crafted and refined complex prompts for LLM testing and evaluation. • Labeled and reviewed multi-turn dialogues between humans and AI for coherence and precision. • Performed RLHF tasks, ranking AI outputs using custom rubrics to reinforce model behavior. • Conducted linguistic evaluations across English, Oromo, and Kiswahili, documenting hallucinations and biases for reliability.

As an AI Data Trainer at Mercor AI, I evaluated and improved large language model outputs to ensure logic, factual accuracy, and helpfulness. I conducted labeling and quality assurance through prompt engineering, multilingual dialogue annotation, and RLHF methodologies. My work enhanced model performance by identifying errors, biases, and supporting expansion to low-resource East African languages. • Crafted and refined complex prompts for LLM testing and evaluation. • Labeled and reviewed multi-turn dialogues between humans and AI for coherence and precision. • Performed RLHF tasks, ranking AI outputs using custom rubrics to reinforce model behavior. • Conducted linguistic evaluations across English, Oromo, and Kiswahili, documenting hallucinations and biases for reliability.

2025 - Present

Education

Work History