Key Skills

Software

No software listed

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Document

Image

Text

Top Label Types

Classification

Object Detection

Question Answering

Text Summarization

Freelancer Overview

I have hands-on experience training and evaluating AI systems as part of a research-focused project in embodied AI using the AI2-THOR simulator. I trained a reinforcement learning–based navigation agent (DDQN) to perform visual object search and goal-directed movement in photorealistic indoor environments. This involved designing state representations, action spaces, reward functions, and training curricula, as well as analyzing training logs to identify failure modes and improve model performance. My work emphasized data quality, consistency, and controlled experimentation during AI training. In addition to navigation training, I developed task-level AI training pipelines that combine reinforcement learning with rule-based interaction control and natural language task execution. I created task generators, task-aware reward systems, and evaluation scripts to measure success across multi-step tasks such as pick-and-place. This experience strengthened my ability to analyze AI behavior, detect edge cases, debug training instabilities, and ensure reliable model evaluation, making me well-suited for AI training, labeling, and quality assurance roles that require precision and attention to detail.

Labeling Experience

Education

Work History