Junde Fong

Key Skills

Software

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Top Label Types

Computer Programming Coding

Function Calling

Prompt Response Writing SFT

Freelancer Overview

I am a full stack engineer with over 7 years of experience building cloud-native and data-intensive platforms, with a strong focus on AI and machine learning solutions. My recent work includes developing an AI-powered crop-monitoring system that leveraged satellite imagery, weather data, and spatial analytics to improve yield and disease prediction accuracy. I have hands-on experience with data preprocessing, feature engineering, and deploying production ML models using TensorFlow and scikit-learn, as well as designing robust data pipelines for real-time and batch inference. My background also includes implementing data quality controls, model performance monitoring, and collaborating closely with data scientists to optimize training data and annotation workflows. I am proficient with tools such as FastAPI, AWS, Azure, and PostGIS, and am passionate about ensuring high-quality, well-labeled data to drive reliable AI outcomes in domains like computer vision and geospatial analytics.

Labeling Experience

Code Human Preference

LabelboxComputer Code ProgrammingComputer Programming CodingFunction Calling

For this task, I should select a codebase that is a git repository, and ask the model to perform a single task in that codebase. The project goal is to, over multiple turns, iterate on the model’s solution for that task until it reaches a “production-ready” state (defined later in these instructions). In addition to iterating on the model’s solution, you should be iterating on the model’s workflow with it to ensure it is working like a real engineer - meaning ensuring the model is reviewing the code it wrote, validating code against task requirements, committing regularly, etc.

For this task, I should select a codebase that is a git repository, and ask the model to perform a single task in that codebase. The project goal is to, over multiple turns, iterate on the model’s solution for that task until it reaches a “production-ready” state (defined later in these instructions). In addition to iterating on the model’s solution, you should be iterating on the model’s workflow with it to ensure it is working like a real engineer - meaning ensuring the model is reviewing the code it wrote, validating code against task requirements, committing regularly, etc.

2025 - 2025

Education

Work History