Lucky Virani

Key Skills

Software

Roboflow

Label Studio

CVAT

Data Annotation Tech

AWS SageMaker

Top Subject Matter

Automatic Number Plate Recognition (ANPR)

Computer Vision

Vehicle Classification

Top Data Types

Image

Text

Computer Code Programming

Top Task Types

Bounding Box

Polygon

Classification

Freelancer Overview

I have hands-on experience annotating computer vision datasets as an AI Data Annotation Specialist at Synergy Telematics Pvt. Ltd., where I annotated 10,000+ images using Roboflow for a production ANPR (Automatic Number Plate Recognition) dashboard. My work involved bounding box annotation for vehicle number plate detection and polygon segmentation for vehicle body classification across 5+ vehicle types — cars, trucks, buses, motorcycles, and auto-rickshaws — along with character-level labeling (A–Z, 0–9) to support OCR model training. I consistently maintained a less than 2% re-annotation error rate, and my annotated datasets were exported in YOLOv5/v8, COCO JSON, and Pascal VOC XML formats and used directly in a deployed fleet telematics system. Beyond annotation, my background as a React.js Developer working on the ANPR dashboard itself gives me a deeper understanding of how training data connects to model performance and real-world AI products. I hold a B.Tech in Computer Science (CGPA 7.78) with coursework in Machine Learning and Computer Vision, and I am comfortable working independently in remote, quality-focused annotation workflows using tools like Roboflow end-to-end — from upload and labeling to review, versioning, augmentation, and export.

Labeling Experience

AI Data Annotation Specialist – Computer Vision

RoboflowImageBounding Box

As an AI Data Annotation Specialist at Synergy Telematics, I annotated over 10,000 images for an ANPR dashboard project focused on vehicle number plate detection and classification. I performed bounding box and polygon segmentation to support computer vision model training and labeled number plate characters for OCR tasks. The datasets were exported in multiple formats including YOLOv5, YOLOv8, COCO JSON, and Pascal VOC XML. • Annotated images covering multiple classes such as cars, motorcycles, trucks, buses, and auto-rickshaws. • Applied bounding box labels for number plate localization and polygon labels for vehicle segmentation. • Labeled alphanumeric number plate characters (A–Z, 0–9) for OCR model development. • Ensured high labeling quality with less than 2% re-annotation error and collaborated with AI/ML engineers to improve guidelines.

As an AI Data Annotation Specialist at Synergy Telematics, I annotated over 10,000 images for an ANPR dashboard project focused on vehicle number plate detection and classification. I performed bounding box and polygon segmentation to support computer vision model training and labeled number plate characters for OCR tasks. The datasets were exported in multiple formats including YOLOv5, YOLOv8, COCO JSON, and Pascal VOC XML. • Annotated images covering multiple classes such as cars, motorcycles, trucks, buses, and auto-rickshaws. • Applied bounding box labels for number plate localization and polygon labels for vehicle segmentation. • Labeled alphanumeric number plate characters (A–Z, 0–9) for OCR model development. • Ensured high labeling quality with less than 2% re-annotation error and collaborated with AI/ML engineers to improve guidelines.

2025 - Present

Education

Work History