Abiodun I.

Key Skills

Software

No software listed

Top Subject Matter

Large language model (LLM) evaluation

RLHF benchmarks

and localization for Yoruba/Nigerian Pidgin

Top Data Types

Text

Document

Top Task Types

RLHF

Fine-tuning

Freelancer Overview

Freelance AI Trainer & Localization Specialist (Remote). Core strengths include N and A. Education includes Master of Science, Nile University of Nigeria (2022) and Bachelor of Science, University of Ibadan (2019). AI-training focus includes data types such as Text and labeling workflows including RLHF and Fine-tuning.

Labeling Experience

Freelance AI Trainer & Localization Specialist (Remote)

TextRLHF

Freelance AI training and evaluation work focused on improving LLM behavior through prompt-based testing and RLHF-style benchmarking. The role includes ranking and assessing model outputs for quality, tone, and safety, alongside translation review to ensure culturally appropriate rendering in Yoruba and Nigerian Pidgin. Fact-checking against authoritative sources is used to reduce hallucinations and improve response reliability. • Craft complex prompts to probe creative, logical, and factual capabilities. • Evaluate and rank LLM outputs using RLHF benchmarks for quality, tone, and safety. • Review and validate AI translations for Yoruba and Pidgin localization nuances. • Perform source-based fact-checking to minimize machine hallucinations.

2025 - Present

Content Writer & Data Annotator (Freelance/Contract)

TextFine-tuning

Contract work generating and annotating structured text datasets used to fine-tune generative AI models. The labeling includes annotating, tagging, and categorizing large datasets to support NLP algorithm training and performance. Additional contributions involve identifying edge cases and documenting potential biases in model-generated responses. • Generate high-quality structured text data for generative AI fine-tuning. • Annotate, tag, and categorize datasets to improve NLP algorithm training. • Identify, document, and report edge cases and biases in AI outputs. • Maintain a consistent accuracy/quality score above 95% across data projects.

2026 - 2026

Education