Kresna Mustika

Key Skills

Software

Appen

Data Annotation Tech

OneForma

Other

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Document

Text

Top Task Types

Audio Recording

Evaluation Rating

Prompt Response Writing SFT

Text Generation

Translation Localization

Freelancer Overview

As an experienced AI Trainer, I possess the ability to accurately classify data based on provided instructions, and to maintain high quality standards on every project. My background in English Education has equipped me with strong language, editing, and transcribing skills, which are crucial for tasks involving linguistic data. I am committed to contributing to the development of AI technology by ensuring optimal data quality and consistency.

Labeling Experience

Break the AI models

OtherTextEvaluation RatingPrompt Response Writing SFT

This project is focused on improving the AI's accuracy by correcting its wrong answers. The primary task is to write prompts that challenge the AI's knowledge and reasoning, causing it to generate factually incorrect or logically flawed responses. Each wrong answer is then carefully evaluated to identify the specific error, and a new, verified response is written to provide a correct and trustworthy example for model training.

This project is focused on improving the AI's accuracy by correcting its wrong answers. The primary task is to write prompts that challenge the AI's knowledge and reasoning, causing it to generate factually incorrect or logically flawed responses. Each wrong answer is then carefully evaluated to identify the specific error, and a new, verified response is written to provide a correct and trustworthy example for model training.

2025 - 2025

Break the AI models

AppenTextEvaluation RatingPrompt Response Writing SFT

This project is focused on improving the AI's accuracy by correcting its wrong answers. The primary task is to write prompts that challenge the AI's knowledge and reasoning, causing it to generate factually incorrect or logically flawed responses. Each wrong answer is then carefully evaluated to identify the specific error, and a new, verified response is written to provide a correct and trustworthy example for model training.

This project is focused on improving the AI's accuracy by correcting its wrong answers. The primary task is to write prompts that challenge the AI's knowledge and reasoning, causing it to generate factually incorrect or logically flawed responses. Each wrong answer is then carefully evaluated to identify the specific error, and a new, verified response is written to provide a correct and trustworthy example for model training.

2025 - 2025

Audio Record, Audio Response Emotion Recognition and Evaluation

Data Annotation TechAudioClassificationEmotion Recognition

This project focuses on creating a comprehensive dataset for audio-based emotion recognition. The tasks include recording audio samples, classifying the primary emotion expressed (e.g., joy, anger, sadness), transcribing and translating the spoken content, and evaluating the intensity and clarity of the emotional expression. The resulting multilingual dataset is designed to train more nuanced and culturally aware AI models

This project focuses on creating a comprehensive dataset for audio-based emotion recognition. The tasks include recording audio samples, classifying the primary emotion expressed (e.g., joy, anger, sadness), transcribing and translating the spoken content, and evaluating the intensity and clarity of the emotional expression. The resulting multilingual dataset is designed to train more nuanced and culturally aware AI models

2025 - 2025

Break the AI models

Data Annotation TechTextEvaluation RatingPrompt Response Writing SFT

This project is focused on improving the AI's accuracy by correcting its wrong answers. The primary task is to write prompts that challenge the AI's knowledge and reasoning, causing it to generate factually incorrect or logically flawed responses. Each wrong answer is then carefully evaluated to identify the specific error, and a new, verified response is written to provide a correct and trustworthy example for model training.

This project is focused on improving the AI's accuracy by correcting its wrong answers. The primary task is to write prompts that challenge the AI's knowledge and reasoning, causing it to generate factually incorrect or logically flawed responses. Each wrong answer is then carefully evaluated to identify the specific error, and a new, verified response is written to provide a correct and trustworthy example for model training.

2025 - 2025

Making a quiz

Data Annotation TechTextQuestion AnsweringText Generation

This project focuses on creating quizzes from source texts across multiple subjects, including Indonesian Language, English Language, Biology, and Chemistry. The core tasks involve Text Generation to create relevant questions and Question Answering to identify the correct answers from the provided context. The final dataset is designed for training AI models in educational applications, with a focus on producing high-quality, relevant, and factually correct quiz content.

This project focuses on creating quizzes from source texts across multiple subjects, including Indonesian Language, English Language, Biology, and Chemistry. The core tasks involve Text Generation to create relevant questions and Question Answering to identify the correct answers from the provided context. The final dataset is designed for training AI models in educational applications, with a focus on producing high-quality, relevant, and factually correct quiz content.

2025 - 2025

Show more

Education

Work History