Silva Std

Key Skills

Software

Data Annotation Tech

Top Subject Matter

No subject matter listed

Top Data Types

Document

Image

Text

Top Task Types

Classification

Evaluation/Rating

RLHF

Text Generation

Translation/Localization

Freelancer Overview

I have hands-on experience as an AI Trainer specializing in both English and Indonesian language data. My work has involved diverse projects such as AI safety classification, natural conversation with chatbots, and providing quality ratings for AI-generated responses. This has equipped me with a sharp sense for linguistic nuance, cultural context, and alignment with AI safety and ethical standards in multiple languages. What sets me apart is my dual-language expertise combined with practical experience in evaluating AI output not just for accuracy, but also for tone, safety, and localization. Additionally, I’ve contributed to peer response evaluations, which honed my critical assessment and consistency in upholding annotation guidelines. With this multidisciplinary approach, I’m confident in my ability to contribute precise, contextually aware data that directly improves AI models’ real-world performance.

Labeling Experience

AI Model Evaluation

Data Annotation TechTextClassificationText Generation

In this project, I was responsible for evaluating AI-generated responses in both English and Indonesian across various tasks, including chatbot conversations, AI safety classifications, and content appropriateness reviews. The project involved assessing thousands of model outputs based on specific quality measures such as accuracy, safety, relevance, tone alignment, and localization. I also contributed by providing peer evaluations, rating fellow trainers' assessments to maintain annotation consistency and data integrity. The dataset consisted of open-domain conversational prompts, classification tasks related to harmful or unsafe content, and multilingual text responses. This project required strong attention to detail, cultural sensitivity, and the ability to make quick, unbiased judgments within tight deadlines while ensuring adherence to project guidelines and quality benchmarks.

In this project, I was responsible for evaluating AI-generated responses in both English and Indonesian across various tasks, including chatbot conversations, AI safety classifications, and content appropriateness reviews. The project involved assessing thousands of model outputs based on specific quality measures such as accuracy, safety, relevance, tone alignment, and localization. I also contributed by providing peer evaluations, rating fellow trainers' assessments to maintain annotation consistency and data integrity. The dataset consisted of open-domain conversational prompts, classification tasks related to harmful or unsafe content, and multilingual text responses. This project required strong attention to detail, cultural sensitivity, and the ability to make quick, unbiased judgments within tight deadlines while ensuring adherence to project guidelines and quality benchmarks.

2024

Education

Work History