Open to opportunities

virginia gichuhi

@virginiagichuhi

Message

AI and language specialist focused on LLM safety, prompt engineering, and high-accuracy data annotation.

Kenya

Message

What I'm looking for

I seek roles where I can apply prompt engineering, LLM training, and annotation expertise in cross-functional teams focused on model safety, cultural alignment, and high-quality dataset development.

I am an AI and language professional with hands-on experience enhancing large language models through prompt engineering, SFT, RLHF, adversarial testing, and red teaming to improve safety, alignment, and performance. I have developed benchmark datasets, ranked outputs for coherence and factual accuracy, and collaborated with engineers and linguists to deliver culturally nuanced, high-quality model behavior.

My background in linguistics and extensive annotation work—transcribing 1,000+ hours at 98%+ accuracy and producing consistent dataset QA—enables me to design robust labeling schemas, refine guidelines, and support evaluation for search, ads, and LLM outputs. I bring a detail-oriented, culturally informed approach to dataset creation, validation, and model evaluation.

Experience

Work history, roles, and key accomplishments

Current

Prompt Engineer

Current

Oneforma

Jan 2024 - Present (1 year 9 months)

Designed and tested prompt templates for summarization, Q&A, and culturally nuanced generation; applied SFT, RLHF, red teaming, and adversarial testing to improve model safety and performance and developed benchmark datasets for evaluation.

Prompt Engineering SFT Red Team Datasets Model Evaluation

Transcriber & Annotator

Rev.com

Jan 2022 - Jan 2024 (2 years)

Transcribed over 1,000 hours of multimedia with 98%+ accuracy, tagged speakers, timestamps and metadata to support ML training, and applied linguistic sensitivity for culturally accurate labeling under tight deadlines.

Transcription Speaker Tagging Metadata QA

Data Annotator

Tech Innovators

Jan 2021 - Jan 2022 (1 year)

Annotated text, image, and audio datasets and conducted QA audits to ensure labeling consistency; refined labeling guidelines, created onboarding documentation, and supported search, ad, and relevance ranking datasets.

Data Annotation Auditing Schema Design Documentation Search Relevance