Key Skills

Software

No software listed

Top Subject Matter

Embedded systems

Python programming

Numerical methods

Top Data Types

Computer Code Programming

Top Task Types

Computer Programming/Coding

Freelancer Overview

I don't have prior commercial experience in AI training data or data labeling, but my master's coursework in machine learning has given me hands-on experience building, training, and evaluating ML models end-to-end — which is the same workflow involved in creating high-quality training and evaluation problems. In one course assignment, I trained and compared non-neural classifiers (SVM and KNN) on a wheat seed dataset using MATLAB, including data shuffling, cross-validation, hyperparameter tuning, hold-out testing, and confusion matrix analysis. In a follow-up assignment, I designed a neural network with Bayesian regularization to estimate bus fuel consumption from a 19-feature dataset, including feature engineering decisions (such as remapping noisy ID columns based on their relationship to the target), architecture and activation function selection, and covariance error evaluation. Both assignments required me to document my reasoning, justify modeling choices, and validate results — closely mirroring the design-and-verify workflow for ML training problems. Beyond ML coursework, my engineering background spans Python (search algorithms, numerical work), C/C++ (embedded systems on Mbed and ESP32), MATLAB/Simulink (control system simulation, including my bachelor's thesis on a real-time PID controller over 5G), and Linux. I'm comfortable framing engineering problems precisely, choosing appropriate computational methods, and verifying numerical answers — which I understand is at the core of designing computational engineering problems for AI evaluation.

Labeling Experience

Education

Work History