Aaron H.

Key Skills

Software

Other

Data Annotation Tech

Scale AI

Top Subject Matter

Multimodal AI (text/image) and next-generation model training

Biology domain for LLM multimodal expert work and RLHF training

Biology/genomics-adjacent scientific competency for multimodal and reasoning LLMs

Top Data Types

Image

Text

Document

Top Task Types

Fine-tuning

RLHF

Question Answering

Freelancer Overview

Ph.D. scientist with deep expertise in biology, genomics, and bioinformatics, currently focused on AI/ML data training and LLM development. Brings 12 years of industry experience in techbio and healthcare software product development, and is an experienced contributor to RLHF workflows for leading frontier models, with roles spanning prompt engineering, response evaluation, rubric development, and multimodal data annotation. Offers a rare combination of scientific domain depth and hands-on AI training experience across text, image, and reasoning model projects. Education includes Doctor of Philosophy, University of California, Berkeley (2013) and National Research Service Award (NRSA) Fellowship, National Institutes of Health (NIH) / University of California, Berkeley (2013). AI-training focus includes data types such as Image, Text, Audio and Videos and labeling workflows including Fine-tuning and RLHF for general multi-modal workflows and biology, STEM and healthcare domain expertise.

Labeling Experience

Multi-Modal Data Trainer (North America Handshake AI)

OtherImageFine-tuning

Responsible for multimodal data annotation and training to support next-generation AI models. Applies dataset preparation and labeling guidance aligned to model training needs. Ensures annotated outputs are suitable for downstream training and evaluation workflows. • Multimodal data annotation and training • Remote North America collaboration • Training data preparation for multimodal models • Model-oriented labeling support

2026 - Present

Biology Domain Expert / LLM Multimodal Expert (North America DataAnnotation)

OtherTextRLHF

Contributes to RLHF workflows for LLM development, including the creation and refinement of training instructions and evaluation artifacts. Participates in prompt engineering and response evaluation to improve model behavior. Develops rubric and reasoning-focused components to support high-quality feedback and training signals. • RLHF prompt development and response evaluation • System instruction development and refinement • Rubric development • CoT/reasoning model support

2025 - Present

Biology Domain Expert / LLM Multimodal Expert (Outlier AI)

OtherTextRLHF

Performs RLHF-driven training for next-generation multimodal and reasoning LLMs targeted to scientific and technical competency. Develops prompts, evaluates responses, and refines system instructions to guide model outputs. Creates rubric and reasoning-oriented elements (e.g., CoT) to standardize feedback and improve training effectiveness. • Reviewer and individual contributor participation • RLHF training for scientific/technical competency • Prompt development and response evaluation • System instruction and rubric development (including CoT/reasoning)

2024 - Present

Education

Work History