Okoegwale E.

Key Skills

Software

No software listed

Top Subject Matter

Africa fintech sector

Regulatory Compliance

Product Research in Africa

Top Data Types

Text

Document

Top Task Types

Classification

Text Generation

Text Summarization

Object Detection

Question Answering

Freelancer Overview

Implement the high false acceptance ratio of End-of-life assets listings on the first e-waste marketplace in Africa, Gocycle to improved accuracy rate of 98%.

Labeling Experience

Image verification

ImageClassification

To reduce the fraudulent listing of end-of-life batteries in a marketplace in Africa by 90%. I trained my model using a transfer learning approach on a custom dataset, gathering thousands of both authentic product pictures and known fraudulent battery listings with altered brand labels and synthetic AI-generated images. Current image accuracy rate is 99% with approval rate of over 1,000 images per hour without human intervention.

2025 - Present

Education

Work History