Ran G.

Key Skills

Software

Other

Top Subject Matter

Retail & Food Service – Demand Forecasting & Inventory Management

Energy & Utilities – Pipeline Monitoring & Leak Detection

FinTech / Financial Services – Risk Indicator System & User Segmentation

Top Data Types

Text

Top Task Types

Bounding Box

Freelancer Overview

I have solid experience in data processing and training data generation. During the "Natural Gas Pipeline Network Leak Detection" project, faced with a severe shortage of real-world leak data, I independently designed a synthetic data generation pipeline using graph theory to produce comprehensive and realistic simulation data. This experience gave me a deep understanding of how training data quality directly impacts model performance, as well as the practical ability to solve data scarcity problems in AI training. Additionally, during my time at Yum China, I led the forecasting project for 15,000 KFC stores across the country, covering the end-to-end pipeline from data cleaning and feature engineering to model development and big data engineering deployment. This further strengthened my ability to process large-scale business data and deliver engineering-ready solutions. My core strengths include: proficiency in Python, PySpark, and Pandas for data cleaning, feature engineering, and statistical analysis; the ability to proactively generate high-quality training samples; and the capability to reason backward from model requirements to define labeling logic and data standards. With hands-on experience in machine learning and deep learning models (e.g., LSTM, BERT for sentiment classification), and a proven track record of 90% accuracy in business-critical forecasting at Yum China, I deeply understand how data format, distribution, and labeling precision affect model performance. I am detail-oriented, responsible, and well-suited for AI training roles that demand high data consistency and quality.

Education

Work History