Aniekan Udo

Key Skills

Software

Appen

Google Cloud Vertex AI

Mindrift

OneForma

Internal/Proprietary Tooling

Top Subject Matter

No subject matter listed

Top Data Types

Document

Image

Text

Video

Top Label Types

Audio Recording

Classification

Computer Programming Coding

Entity Ner Classification

Evaluation Rating

Object Detection

Freelancer Overview

I am an AI engineer with hands-on experience designing, evaluating, and improving large language models (LLMs) through rigorous data labeling, annotation, and automated evaluation frameworks. My work has focused on prompt engineering, output scoring, and reducing hallucinations, achieving a 40% drop in AI hallucinations and 97% answer accuracy on production retrieval-augmented generation (RAG) systems. I have built and integrated prompt regression testing with over 50 automated test cases, developed RLHF-style feedback loops, and implemented quality scoring agents that significantly reduced manual review time. My technical skills span Python, LangChain, LangGraph, OpenAI, Anthropic, and data pipeline tools like PostgreSQL and MongoDB. I am passionate about ensuring high-quality, reliable AI outputs through structured data annotation, continuous evaluation, and process automation, making me well-suited for roles focused on AI training data and output quality.

Labeling Experience

data annotator

MindriftTextClassificationEvaluation Rating

Evaluated pair-wise search engine results to determine the superior option, contributing to refined algorithms and more effective recommendation systems. Reviewed products on shopping websites, ensuring accurate category alignment and improving overall data integrity. Compared AI-generated responses for safety, helpfulness, and factual accuracy, identifying areas for enhancement.

Evaluated pair-wise search engine results to determine the superior option, contributing to refined algorithms and more effective recommendation systems. Reviewed products on shopping websites, ensuring accurate category alignment and improving overall data integrity. Compared AI-generated responses for safety, helpfulness, and factual accuracy, identifying areas for enhancement.

2025 - 2025

Education

Work History