Ari Kurniawan Saputra

Key Skills

Software

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Text

Top Task Types

Audio Recording

Text Generation

Freelancer Overview

I have practical experience contributing to AI training and evaluation through various data annotation and interaction-based projects. These include assessing AI assistants in both text-based and spoken scenarios, where I evaluated their ability to perform technical tasks, handle natural conversations, and generate accurate responses. My work involved interacting with AI systems using real-world tools (such as cloud platforms), recording conversational data, and rating AI-generated replies based on quality and relevance.

Labeling Experience

Visual Environment Interaction (RealTime Pocket Assistant V2 Project)

LabelboxVideoQuestion AnsweringEvaluation Rating

Recorded 3–5 minute mobile videos while performing everyday activities (e.g., cooking, shopping, walking outside) and interacting with an AI video assistant. The task involved prompting the AI to describe visual elements, provide feedback, or assist with tasks based on real-time camera input and conversation.

Recorded 3–5 minute mobile videos while performing everyday activities (e.g., cooking, shopping, walking outside) and interacting with an AI video assistant. The task involved prompting the AI to describe visual elements, provide feedback, or assist with tasks based on real-time camera input and conversation.

2025 - 2025

AI Response Evaluation & Annotation (Project Omni Indonesian v2)

LabelboxTextEvaluation RatingPrompt Response Writing SFT

Performed quality evaluation of AI-generated responses in Indonesian by reviewing full conversation histories. Applied detailed annotation guidelines to compare, rate, and provide structured feedback on AI outputs based on relevance, coherence, and adherence to task instructions.

Performed quality evaluation of AI-generated responses in Indonesian by reviewing full conversation histories. Applied detailed annotation guidelines to compare, rate, and provide structured feedback on AI outputs based on relevance, coherence, and adherence to task instructions.

2025 - 2025

Conversational AI Testing (Real Voices Indonesian Project)

LabelboxAudioEvaluation RatingData Collection

AI stress-testing through natural spoken interactions under varied acoustic conditions. Generated high-quality audio data in Indonesian to assess AI conversational robustness in realistic scenarios.

AI stress-testing through natural spoken interactions under varied acoustic conditions. Generated high-quality audio data in Indonesian to assess AI conversational robustness in realistic scenarios.

2025 - 2025

Interactive AI Evaluation (ScreenCopilot Project)

LabelboxVideoAction RecognitionEvaluation Rating

Recorded and annotated interactive screen recordings involving technical software tools (e.g., AWS, GCP, Salesforce), while prompting an AI assistant to solve technical tasks. This involved evaluating the AI’s contextual understanding and task-handling ability in real-time environments.

Recorded and annotated interactive screen recordings involving technical software tools (e.g., AWS, GCP, Salesforce), while prompting an AI assistant to solve technical tasks. This involved evaluating the AI’s contextual understanding and task-handling ability in real-time environments.

2025 - 2025

Education

Work History