Oluwabi M.

Key Skills

Software

Don't disclose

Top Subject Matter

Electrical engineering diagrams and financial chart analysis

Legal Services & Contract Review

Regulatory Compliance & Risk Analysis

Top Data Types

Text

Document

Top Task Types

Tracking

Freelancer Overview

Multimodal AI Evaluation Project (Independent Contractor). Brings 9+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Don't disclose. Education includes Master of Business Administration, Ahmadu Bello University (2025) and Bachelor of Science, Olabisi Onabanjo University (2016). AI-training focus includes data types such as Text and labeling workflows including Evaluation, Rating, and Tracking.

Labeling Experience

Multimodal AI Evaluation Consultant - Independent Contractor

TextTracking

Evaluated multimodal LLM outputs that included electrical diagrams and financial charts for accuracy and reasoning quality. Identified technical inaccuracies and reasoning gaps by cross-checking diagram and financial context. Developed structured evaluation frameworks and prompt-testing methodologies to improve model performance. • Review and assess model outputs involving technical diagrams and financial data • Flag discrepancies, inaccuracies, and reasoning gaps during evaluations • Create evaluation rubrics and structured scoring frameworks • Test and refine prompts using repeatable prompt-testing methods

2025 - Present

Multimodal AI Evaluation Project (Independent Contractor)

Don't discloseText

Evaluated multimodal LLM outputs that include electrical diagrams and financial charts to assess correctness. Identified technical inaccuracies and reasoning gaps and provided structured feedback to improve model accuracy. Built evaluation frameworks and prompt-testing methodologies to support consistent quality measurement across outputs. • Reviewed electrical diagram-related model outputs for technical validity • Checked financial chart-related outputs for consistency and correctness • Documented structured observations to highlight error types and reasoning gaps • Designed repeatable prompt tests to compare model performance

2025 - Present

Education

Work History