Safira Gita Nabilla

Key Skills

Software

Appen

Clickworker

CrowdSource

CVAT

Google Cloud Vertex AI

Lionbridge

Mighty AI

Mindrift

OneForma

Remotasks

Toloka

Telus

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Video

Top Task Types

Classification

Data Collection

Evaluation Rating

Object Detection

Translation Localization

Freelancer Overview

An expert linguist and data annotator, I’ve contributed to numerous AI training projects over the past 6+ years with a focus on LLM evaluation, query annotation, and multimedia classification. My expertise spans a wide range of data types including text, image, video, and audio. I specialize in high-accuracy labeling tasks such as evaluation and rating, data collection, object detection, localization, and attribute classification. With an academic background in Language and Literature, my skills align closely with language-driven AI tasks. I’m confident using a variety of annotation tools provided by clients and quickly adapt to new platforms and workflows. Some of the key projects I’ve worked on including LLM evaluation in Javanese and Indonesian, Prompt evaluation and rating, Health integrity labeling, Photo RCV attribute annotation, Query creation, validation, and labeling, Ads quality rating, etc.

Labeling Experience

Photo RCV Attribute Annotation

TelusImageClassificationObject Detection

The project aims to label skin tone, age, and gender in photos and videos to evaluate computer vision model performance and identify potential biases. The dataset includes hundreds of thousands of images and videos, with a target labeling accuracy of 95%.

The project aims to label skin tone, age, and gender in photos and videos to evaluate computer vision model performance and identify potential biases. The dataset includes hundreds of thousands of images and videos, with a target labeling accuracy of 95%.

2021 - 2022

Health Integrity Topic Labelling

LionbridgeVideoSegmentationClassification

Reviewed and labeled social media posts (photo, text, video, or link) by classifying topics (health, human vaccine, COVID, flu, masks), identifying whether content was COVID vaccine-related or not, assigning sentiment (positive, neutral, negative), and selecting relevant COVID-19 vaccine safety concerns. Maintained over 95% accuracy across 500–1000 tasks/day, within a high-volume project of tens of thousands of entries replenished continuously over a 2-year period.

Reviewed and labeled social media posts (photo, text, video, or link) by classifying topics (health, human vaccine, COVID, flu, masks), identifying whether content was COVID vaccine-related or not, assigning sentiment (positive, neutral, negative), and selecting relevant COVID-19 vaccine safety concerns. Maintained over 95% accuracy across 500–1000 tasks/day, within a high-volume project of tens of thousands of entries replenished continuously over a 2-year period.

2021 - 2022

Education

Work History