Fanzhi S.

Key Skills

Software

No software listed

Top Subject Matter

Physics education and expert scientific reasoning

Materials science education and AI training content

Electron microscopy image segmentation for battery electrode analysis

Top Data Types

Text

Image

Document

Top Task Types

Question Answering

Segmentation

Freelancer Overview

SepalAI Remote — Physics Researcher (AI Training & Content Development). Brings 6+ 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 Doctor of Philosophy, The University of Cambridge (2021) and Master of Philosophy, The University of Cambridge (2021). AI-training focus includes data types such as Text and Image and labeling workflows including Evaluation, Rating, and Question Answering.

Labeling Experience

University of Cambridge — Graduate Researcher (Machine Learning for Nano and Device Materials Electron Microscopy Signal Analysis)

ImageSegmentation

Developed large-scale paired electron microscopy datasets and established standardized preprocessing, labeling, and quality control pipelines. Implemented end-to-end deep learning workflows for automated image reconstruction and analysis, including defect detection and denoising. Applied evaluation metrics and segmentation-quality improvements (e.g., patch-based blending, PSNR/SSIM, perceptual loss) to robustly enhance reconstructed image fidelity. •Built and curated 22,000+ paired electron microscopy image dataset.•Performed data preprocessing, labeling, and quality control in standardized pipelines.•Developed automated microscopy reconstruction and defect detection models.•Improved output quality using PSNR/SSIM and perceptual loss with patch-based blending.

2021 - Present

SepalAI Remote — Physics Researcher (AI Training & Content Development)

TextQuestion Answering

Created advanced materials-science problems and comprehensive marking schemes to serve as high-quality training data for AI projects. Ensured content met rigorous scientific accuracy and pedagogical standards prior to quality review. Focused on producing expert-level question content suitable for AI tutoring and training pipelines.•Designed and developed 50+ advanced materials science problems.•Authored comprehensive marking schemes for three AI training projects.•Passed content through quality review criteria.•Maintained rigorous scientific and pedagogical standards.

2024 - 2026

SepalAI Remote — Physics Researcher (AI Training & Content Development)

Don't discloseText

Contributed to AI training and content development by reviewing and evaluating physics-olympiad-level questions for scientific depth and reasoning validity. Used expert physics knowledge to assess alignment with competition-standard benchmarks to support model calibration. Helped improve how large language models reason about physical systems and phenomena.•Reviewed 30+ physics-olympiad-level questions.•Assessed reasoning validity and scientific depth.•Verified benchmark alignment for expert-level calibration.•Evaluated AI-generated scientific outputs for accuracy and clarity.

2024 - 2026

DP Technology Beijing, China — Computer Vision Assisted Battery Electrode Electron Microscopy Analysis

OtherImageSegmentation

Built and annotated a dataset of electron microscopy images to support computer-vision model training for secondary particle segmentation. Used SAM labelling tools and COCO-format viewers to create training labels for downstream deep-learning workflows. Delivered segmentation-focused analysis that improved materials characterization via image-based statistical assessment. •Annotated 200+ electron microscopy images for segmentation.•Prepared dataset and labeling pipeline using SAM labelling tool and COCOViewer.•Supported model training for precise secondary particle segmentation.•Applied image-based statistical analysis for improved assessment.

2023 - 2023

Education

Work History