Rhobert K.

Key Skills

Software

OneForma

Don't disclose

Other

Scale AI

Top Subject Matter

Generative AI model training

LLM evaluation

Alignment Domain Expertise

Top Data Types

Text

Audio

Image

Document

Top Task Types

RLHF

Transcription

Freelancer Overview

GenAI Model Trainer & AI Evaluation Specialist | OneForma (Remote). Core strengths include OneForma, Don't disclose, and Other. Education includes Bachelor of Science, University of Nairobi (2019). AI-training focus includes data types such as Text and Audio and labeling workflows including RLHF, Evaluation, and Rating.

Labeling Experience

GenAI Model Trainer & AI Evaluation Specialist | OneForma (Remote)

OneFormaTextRLHF

Trained and evaluated generative AI model outputs using RLHF-oriented preference and comparative ranking workflows. Assessed responses for helpfulness, factual accuracy, safety, instruction adherence, reasoning quality, and bias presence according to provided guidelines and benchmarks. Provided evaluator feedback that was used to improve alignment and future model training and optimization. • Ranked LLMs by multi-criteria evaluations (helpfulness, factual accuracy, safety, instruction adherence, reasoning, tone). • Performed preference ranking and comparative analysis of response pairs for reinforcement learning tasks. • Detected hallucinations, logical fallacies, edge cases, and biases using advanced reasoning checks. • Delivered structured evaluator feedback to support iterative model training and optimization.

2024 - 2026

Search Specialist & AI Evaluator | TELUS International AI/UHRS (Remote)

Don't discloseText

Performed search relevance assessments, including query-relevance matching, query intent satisfaction, landing page evaluation, and ranking judgments. Evaluated online and social media content for risk of misinformation, policy violations, toxicity, and sentiment using strict guideline calibration. Achieved high benchmark agreement scores while executing QA checklist and edge-case based reduction of rejection rates. • Conducted query-relevance matching and intent satisfaction labeling across search results. • Assessed landing page quality and ranking relevance for search evaluation tasks. • Moderated and rated content for misinformation risk, policy compliance, toxicity, and sentiment. • Used calibration and QA checklists to reduce rejection rates by 20%.

2022 - 2023

Data Annotator | Scale AI (Remote)

Scale AIAudioTranscription

Labeled and reviewed audio datasets including ASR transcription and speaker diarization tagging for intent classification tasks. Completed computer vision-style annotations on difficult datasets, including segmentation, polygons, keypoints, and bounding boxes. Met gold-standard task accuracy above 98% while improving annotation throughput and reducing downstream QA rejections through validation and review procedures. • Produced ASR transcriptions and speaker diarization labels for audio recordings. • Labeled intent classification outputs associated with the audio content. • Performed bounding box, polygon/segmentation, and keypoint annotations on challenging data. • Maintained high-quality KPIs, achieving 10–20% above weekly goals and reducing QA rejections by 25%.

2021 - 2022

AI Data Annotator & Content Evaluator | RWS/Aligner Contract (Remote)

OtherText

Completed training and annotation for text evaluations, response ranking, and quality assessment tasks using comprehensive guidelines and benchmarks. Assessed AI-generated content for compliance with content safety, policy requirements, and datavalidation checks. Supported distributed team processes to ensure annotation accuracy and project performance. • Labeled and evaluated text responses using benchmark criteria and rubric-based guidelines. • Performed content moderation and content safety assessments for policy compliance. • Conducted data validation activities tied to annotation quality. • Collaborated with distributed teams to maintain accuracy and consistency.

2020 - 2021

Education

Work History