Open to opportunities

Ondira Naboth

@ondiranaboth

Message

Linguistic expert and AI data evaluator specializing in Swahili and English model safety.

Kenya

Message

What I'm looking for

I seek roles focused on LLM safety and evaluation, adversarial testing, and multilingual dataset validation within collaborative teams that prioritize robust, ethical AI and opportunities for technical growth.

I am an AI data evaluator and linguistic specialist with hands-on experience annotating, reviewing, and stress-testing multilingual (Swahili & English) datasets and large language models. I focus on safety evaluation, adversarial prompt design, edge-case discovery, policy compliance, and model robustness, bringing a technical background in information systems, AI fundamentals, and cybersecurity to structured, specification-driven evaluation work.

My contract roles include adversarial prompt design and red teaming, Swahili and English dataset evaluation, LLM operations and evaluation, and cybersecurity monitoring. I document failure cases, provide mitigation feedback, validate localized content for AI training readiness, and maintain evaluation pipelines and structured QA workflows to improve model alignment and output quality.

Experience

Work history, roles, and key accomplishments

Current

AI Red Teamer

Current

Independent

Sep 2025 - Present (6 months)

Designed adversarial prompts to test model robustness and safety boundaries, identified jailbreak patterns and hallucinations, and documented failure cases with structured mitigation feedback.

Adversarial Prompting Model Safety Testing Jailbreak Detection LLM Evaluation Risk Analysis & QA

Current

AI Language Data Evaluator

Current

Independent

Jan 2024 - Present (2 years 2 months)

Evaluated Swahili and English datasets for semantic accuracy, intent alignment, and tone preservation, conducted QA checks to identify edge cases and guideline deviations, and validated localized content for AI training readiness.

Data Annotation Sentiment Analysis Localization NER Edge Cases

LLM Operations Specialist

Zone1 Kenya

Oct 2024 - Oct 2025 (1 year)

Supported deployment and monitoring of large language models, assisted evaluation pipelines for model output quality and reliability, and maintained structured workflows for dataset versioning and evaluation feedback.

Evaluation Pipelines Reliability Operations Support

Cybersecurity Intern

National Cereals Board

Jan 2024 - Aug 2024 (7 months)

Monitored system logs and alerts to identify anomalies, assisted in vulnerability assessments and compliance checks, and applied procedural accuracy transferable to data quality and AI evaluation work.

Vulnerability Assessment & Compliance