Abhigyan D.

Key Skills

Software

Scale AI

Don't disclose

Other

Top Subject Matter

Frontier LLM evaluation and reward-signal grading (RLHF-adjacent) for technical/quantitative document artifacts

Frontier-model evaluation and RLHF training data for technical and quantitative domains

ML training data annotation

Top Data Types

Text

Document

Top Task Types

RLHF

Classification

Freelancer Overview

AI Output Rater / Model Evaluation Specialist — Feather. Brings 3+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Don't disclose, Internal, and Proprietary Tooling. Education includes Bachelor of Science, University of Toronto (2022). AI-training focus includes data types such as Document and Text and labeling workflows including RLHF, Evaluation, and Rating.

Labeling Experience

AI Output Rater / Model Evaluation Specialist — Feather

Don't discloseDocumentRLHF

Provided human-feedback evaluation for frontier-model training using detailed HL grading rubrics. Rated AI-generated slide-deck artifacts for factual accuracy, instruction-following, design quality, and structural coherence to generate reward-relevant signals. Applied quantitative and technical domain knowledge to assess complex multi-step outputs where general raters may lack correctness context. • Evaluated outputs for the “Les Artistes — Artifacts” HL Grading campaign • Produced high-precision, granular feedback consistent across large batches • Graded technical content, document structure, and reasoning depth • Maintained rubric adherence for training pipeline inputs

2025 - Present

AI Labeller

Scale AITextRLHF

Architected and reviewed large-scale Python/SQL ML pipelines serving 10M+ daily requests, achieving 99.5% uptime through comprehensive testing and optimization • Reduced production ML system runtime by 40% through performance profiling, debugging, and infrastructure improvements, enhancing deployment reliability • Enhanced LLM performance by 35% using advanced RLHF optimization, prompt engineering, and A/B testing frameworks • Led cross-functional initiatives with 5+ engineering teams, implementing auto-feedback loops that improved system efficiency by 28% • Collaborated with analytics and monitoring teams to present model insights and technical concepts to nontechnical stakeholders.

2025 - 2025

AI / ML Engineer — Outlier AI

Document

Contributed to frontier-model evaluation and RLHF-adjacent training data projects through prompt and response assessment. Performed response ranking and correctness assessment on technical and quantitative tasks to support training-data quality. Authored expert reasoning traces for math, statistics, and programming to improve evaluative supervision signals. • Conducted prompt evaluation and response ranking • Assessed correctness for technical/quantitative tasks • Produced expert reasoning traces for supervision • Supported large-scale model evaluation and training-data workflows

2024 - 2025

Data Analyst & ML Engineer — Crowdgen

OtherTextClassification

Performed structured data annotation and quality assurance to support ML training workflows. Conducted output evaluation as part of model-feedback pipelines to help improve training data reliability. Implemented data-processing scripts to enforce higher inter-rater consistency and data quality. • Structured data labeling tasks • Annotation QA and reliability checks • Output evaluation for feedback pipelines • Data-processing automation for quality control

2023 - 2024

Education

Work History