Abdulganiyu L.

Key Skills

Software

No software listed

Top Subject Matter

AI content generation quality and safety checking for a digital writing tool

AI image generation verification for web design assets

Prompt testing and model behavior evaluation for instruction-following and safety

Top Data Types

Text

Image

Top Task Types

RLHF

Freelancer Overview

AI Content Checker (SabiWriter Project). Brings 2+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include ChatGPT and Google Gemini. AI-training focus includes data types such as Text and Image and labeling workflows including Evaluation, Rating, and RLHF.

Labeling Experience

Daily AI Tester (Personal Practice)

TextRLHF

Tested text and image generation behavior by experimenting with different prompt formulations and instruction constraints. Tracked failure cases such as complete disregard of rules (e.g., not including people) and mistakes inside generated images. Compared instruction-following, quality, and safety differences between ChatGPT and Gemini over time.• Ran weekly experiments to probe model limits.• Documented when models ignored prohibitions or created errors.• Assessed how lighting, camera angles, and objects were handled.• Monitored quality and safety rule behavior across two AI tools.

2025 - Present

Image and Layout Reviewer (Personal Web Projects)

Image

Generated images and checked each output for compliance with detailed style and layout instructions. Verified that incremental edits to a single part of an image did not unintentionally alter other regions. Rejected and regenerated images that were blurry, confusing, or poorly produced.• Reviewed colors and visual details against exact prompts.• Performed careful verification after small image modifications.• Iterated on image outputs by redoing unacceptable results.• Produced icons and layouts for personal web projects using AI generation.

2025 - Present

AI Content Checker (SabiWriter Project)

Text

Reviewed AI-generated text for rule adherence and safety before publishing. Evaluated responses by grading quality and identifying when the model ignored instructions or added unsafe content. Ensured the final output matched exactly what the AI was told to write.• Managed daily review cycles for an AI content writing tool.• Deleted drafts where rules were violated or unsafe material appeared.• Wrote and tuned hundreds of prompts to elicit reliable results.• Prepared AI text for end users by confirming it was safe and readable.

2025 - Present

Education

Work History