Baron J.

Key Skills

Software

Scale AI

Telus

Other

Top Subject Matter

Large language model evaluation

RLHF workflows

AI safety

Top Data Types

Text

Document

Top Task Types

Classification

RLHF

Red Teaming

Freelancer Overview

Senior AI Trainer & Language Specialist at Scale AI. Brings 10+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Scale AI, Telus, and LanguageTool. Education includes Master of Arts, University of Port Harcourt (2022) and Bachelor of Arts, University of Port Harcourt (2020). AI-training focus includes data types such as Text and labeling workflows including Evaluation, Rating, and Classification.

Labeling Experience

Senior AI Trainer & Language Specialist at Scale AI

Scale AIText

Led large-scale AI evaluation and human feedback programs to improve large language model quality and efficiency across multiple domains. Evaluated tens of thousands of AI-generated responses using structured rubrics for reasoning, writing, coding, factuality, and instruction-following. Performed safety and bias verification through red-team testing and hallucination pattern analysis. • Evaluated 25,000+ AI-generated responses and provided consistent reviewer scoring using rubric-driven guidance • Developed evaluation benchmark datasets to support model alignment testing • Increased reviewer consistency by 34% through rubric design and calibration • Trained and mentored 100+ freelance evaluators for annotation/evaluation governance

2024 - Present

Lead AI Content Evaluator at TELUS Digital

TelusText

Managed AI content quality assessment and evaluation programs for conversational outputs across languages and content teams. Reviewed extremely large volumes of AI-generated text using quality accuracy targets and ranking/evaluation frameworks. Drove reviewer training material adoption and reduced annotation disagreement through quality audits and governance. • Reviewed 15 million+ words of AI-generated content while maintaining 98%+ quality accuracy • Built evaluation frameworks to rank conversational AI outputs • Reduced annotation disagreement rates by 27% via audits and standardized reviewer guidance • Led cross-team quality audits and created reviewer training materials for global operations

2022 - 2024

Large Language Model Evaluation Program at ProWritingAid

TextRLHF

Created LLM evaluation guidelines and supervised human evaluation processes to support reinforcement learning from human feedback workflows. Ranked thousands of model outputs for creative and technical writing tasks and documented systematic reasoning failures to guide improvement efforts. Improved instruction adherence and response scoring through iterative prompt refinement and evaluation. • Developed evaluation guidelines adopted by 50+ human annotators • Ranked 10,000+ model outputs to optimize RLHF processes • Identified and documented 200+ reasoning failures for targeted fine-tuning • Improved instruction adherence metrics by 22% using systematic prompt refinement and response scoring

2019 - 2022

Content Quality Manager at LanguageTool

Text

Oversaw linguistic quality assurance and evaluation processes for machine-learning driven writing suggestions and grammar systems. Conducted evaluation of language quality models and built grammar evaluation frameworks for ML teams and enterprise benchmarking. Reduced false-positive grammar alerts and coordinated with computational linguists to improve the accuracy of suggestions. • Evaluated language quality models used by millions of users • Built grammar evaluation frameworks for machine-learning teams • Reduced false-positive grammar alerts by 22% through updated editorial review processes • Designed style evaluation benchmarks for enterprise customers in collaboration with linguists

2017 - 2019

Senior Editor & Fact-Checking Specialist at TheCable

OtherTextClassification

Performed editorial review and evidence-based fact verification for research publications and reports. Checked sources against claims to maintain documented factual accuracy standards across editorial outputs. Trained junior editors on verification processes to improve consistency in the fact-checking workflow. • Fact-checked 3,000+ articles, reports, and publications • Maintained 99%+ documented factual accuracy rate • Developed source-verification methodologies used across editorial teams • Trained junior editors in evidence-based verification for higher-quality outputs

2015 - 2017

Show more

Education

Work History