Sayan Ghosal

AI Data Annotation Specialist | Robotics AI Evaluation

India, IN.

About

Highly accomplished AI Data Annotation Specialist with over 4 years of experience, specializing in LiDAR 3D, Computer Vision, and Video Annotation, alongside Robotics AI Evaluation. Proven ability to deliver high-quality, reliable training data for machine learning models, consistently achieving 98%+ annotation accuracy across diverse and large-scale autonomous vehicle datasets. Expert in identifying performance issues, conducting rigorous quality reviews, and optimizing dataset validation to significantly improve AI model reliability and performance.

Work

Awign

|

AI Video Annotator (Robotics AI Evaluation)

India, , India

→

Summary

As an AI Video Annotator, I evaluated AI-powered household robot task execution videos to enhance task completion accuracy and overall performance.

Highlights

Systematically evaluated AI-powered household robot task execution videos, rigorously assessing task completion accuracy and overall performance against predefined guidelines.

Identified and categorized behavioral errors, task failures, and execution inconsistencies in robot interactions with real-world environments, providing detailed feedback on decision-making and object handling.

Generated high-quality evaluation data and structured annotations, directly contributing to the improvement of AI model reliability and continuous learning for robotics systems.

Maintained consistent annotation accuracy and data integrity while efficiently processing large volumes of robotics and AI-generated video data.

Collaborated effectively with project teams to ensure strict adherence to evaluation standards and annotation guidelines, optimizing data quality and consistency.

Imerit Technology Services Pvt. Ltd.

|

Data Annotator (LIDAR / 3D / Computer Vision)

India, , India

→

Summary

As a Data Annotator, I delivered high-quality LiDAR and image annotations for autonomous vehicle datasets, achieving 98%+ accuracy across large-scale projects.

Highlights

Delivered high-quality LiDAR and image annotations for complex autonomous vehicle datasets, consistently maintaining 98%+ annotation accuracy across large-scale projects.

Expertly annotated intricate 3D point cloud data using advanced techniques like cuboids, bounding boxes, polygons, and polylines to enhance object detection and scene understanding.

Achieved and consistently exceeded productivity targets of 500+ frames per day while performing critical tasks such as lane marking, drivable area segmentation, and object labeling.

Conducted rigorous Quality Check (QC) reviews to identify and rectify annotation errors, significantly improving dataset consistency and reducing error rates.

Collaborated cross-functionally to streamline annotation workflows, adapt to diverse client requirements, and ensure timely project delivery across multiple platforms.

Dulichand Motors Pvt Ltd

|

Back Office Executive

India, , India

→

Summary

As a Back Office Executive, I managed vehicle loan coordination and back office operations, ensuring accurate record maintenance.

Highlights

Streamlined vehicle loan coordination processes, enhancing efficiency in financial transactions and customer service.

Managed comprehensive back office operations, ensuring meticulous maintenance of accurate records and documentation.

Contributed to operational efficiency by overseeing administrative tasks and supporting sales processes.

Binayak Automobiles and Infrastructure

|

Back Office Executive

India, , India

→

Summary

As a Back Office Executive, I efficiently handled insurance, sales tracking, and registration processes for automobile operations.

Highlights

Managed critical insurance documentation and claims processing, ensuring compliance and customer satisfaction.

Implemented robust sales tracking systems, providing accurate data for performance analysis and strategic planning.

Facilitated seamless vehicle registration processes, ensuring timely completion and adherence to regulatory requirements.

Education

Netaji Subhash Open University

India, , India
→
Bachelor of Arts (BA) Arts

Mathurapur Arya Vidyapith HS

India, , India
→
Higher Secondary Higher Secondary

Ballavpur Mohini Mohan High School

India, , India
→
Secondary Secondary

Languages

English

Bengali

Hindi

Skills

Polygon Annotation

Polygon Annotation.

Polyline Annotation

Polyline Annotation.

LIDAR 3D Annotation

LIDAR 3D Annotation.

Point Cloud Annotation

Point Cloud Annotation.

Computer Vision Data Annotation

Computer Vision Data Annotation.

3D Cuboids

3D Cuboids.

Bounding Box

Bounding Box.

Lane Marking & Drivable Area Segmentation

Lane Marking & Drivable Area Segmentation.

Object Detection & Image Annotation

Object Detection, Image Annotation.

Video Annotation

Video Annotation.

AI Model Evaluation

AI Model Evaluation.

Robotics AI Evaluation

Robotics AI Evaluation.

Quality Check (QC) & Review

Quality Check, QC, Review.

Dataset Validation & Error Reduction

Dataset Validation, Error Reduction.

Large-scale Data Handling

Large-scale Data Handling.

Scale AI

Scale AI, LiDAR Annotation, Image Annotation.

iMerit Annotation Tool

iMerit Annotation Tool.

Client Proprietary Annotation Tools

Client Proprietary Annotation Tools.

Google Sheets

Google Sheets, Data Tracking, QC Reporting.

Microsoft Excel & Word

Microsoft Excel, Microsoft Word.

Canva

Canva, Basic Design, Documentation Support.

Projects

Robotics AI Video Evaluation Project

→

Summary

Evaluated AI-powered household robot task execution videos to identify performance gaps, validate task completion, and provide structured feedback for model improvement.

Autonomous Vehicle LiDAR Annotation Project

→

Summary

Worked on large-scale LiDAR and image datasets for autonomous vehicle model training, focusing on object detection, lane marking, and drivable area segmentation.

3D Point Cloud Annotation

→

Summary

Performed detailed annotation of 3D LIDAR data using cuboids and segmentation techniques to improve object recognition accuracy.

Lane & Drivable Area Segmentation

→

Summary

Annotated road scenes to identify lane boundaries and drivable areas, supporting real-world driving scenario modeling.