Vinod Latwal

Key Skills

Software

Label Studio

Roboflow

Top Subject Matter

Natural Language Processing

Rlhf Domain Expertise

Model Evaluation

Top Data Types

Text

Image

Video

Top Task Types

Entity Ner Classification

Question Answering

Classification

Freelancer Overview

AI Data Annotation & Training Data Contributor. Brings 5+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Label Studio and Roboflow. Education includes Bachelor of Commerce, Aligarh Muslim University (2021). AI-training focus includes data types such as Text, Image, and Video and labeling workflows including Entity (NER) Classification, Evaluation, and Rating.

Labeling Experience

AI YouTube Thumbnail Generator Data Annotator

RoboflowImage

For the AI YouTube Thumbnail Generator project, I curated and annotated prompt-response datasets to guide generative AI output. This involved labeling and validating AI-generated image outputs for relevance and quality in content moderation workflows. I also maintained structured data storage pipelines for downstream model evaluation and review. • Structured and labeled title, style, and color data for prompt evaluation. • Performed content quality reviews on AI-generated thumbnails for user intent alignment. • Managed metadata schemas in MongoDB for labeled asset tracking. • Applied real-world evaluation techniques to ensure high-quality output data.

For the AI YouTube Thumbnail Generator project, I curated and annotated prompt-response datasets to guide generative AI output. This involved labeling and validating AI-generated image outputs for relevance and quality in content moderation workflows. I also maintained structured data storage pipelines for downstream model evaluation and review. • Structured and labeled title, style, and color data for prompt evaluation. • Performed content quality reviews on AI-generated thumbnails for user intent alignment. • Managed metadata schemas in MongoDB for labeled asset tracking. • Applied real-world evaluation techniques to ensure high-quality output data.

2026 - 2026

Video Conferencing App Data Annotator

Label StudioVideoClassification

For the Full-Stack Video Conferencing App project, I collected and labeled user interaction data from live video sessions to support behavioral model training. I annotated video/audio metadata for speaker identification, sentiment analysis, and session summarization. My work contributed to developing analytics for performance evaluation and interview scheduling features. • Labeled interaction and engagement signals across multiple video sessions. • Annotated video and audio metadata for behavioral analytics models. • Supported features such as speaker ID, sentiment detection, and summaries. • Contributed to model training for session understanding and content analysis.

For the Full-Stack Video Conferencing App project, I collected and labeled user interaction data from live video sessions to support behavioral model training. I annotated video/audio metadata for speaker identification, sentiment analysis, and session summarization. My work contributed to developing analytics for performance evaluation and interview scheduling features. • Labeled interaction and engagement signals across multiple video sessions. • Annotated video and audio metadata for behavioral analytics models. • Supported features such as speaker ID, sentiment detection, and summaries. • Contributed to model training for session understanding and content analysis.

2025 - 2025

AI Mock Interview Platform Annotator

Label StudioTextQuestion Answering

On the AI Mock Interview Preparation Platform, I designed and annotated adaptive question-answer datasets for interview simulations. My role included careful annotation across varied technical domains and scoring AI-generated responses using defined rubrics. I developed analytics dashboards to track annotation accuracy and feedback for calibration. • Built datasets with domain-specific annotated Q&A pairs. • Evaluated AI-simulated interview responses and provided rubric-based ratings. • Supported model improvement through feedback loop iteration. • Tracked feedback metrics to calibrate data quality standards.

On the AI Mock Interview Preparation Platform, I designed and annotated adaptive question-answer datasets for interview simulations. My role included careful annotation across varied technical domains and scoring AI-generated responses using defined rubrics. I developed analytics dashboards to track annotation accuracy and feedback for calibration. • Built datasets with domain-specific annotated Q&A pairs. • Evaluated AI-simulated interview responses and provided rubric-based ratings. • Supported model improvement through feedback loop iteration. • Tracked feedback metrics to calibrate data quality standards.

2025 - 2025

AI Data Annotation & Training Data Contributor

Label StudioTextEntity Ner Classification

As an AI Data Annotation & Training Data Contributor at DXC Technology, I labeled and annotated structured and unstructured text datasets to support machine learning model development. My work included entity recognition, intent classification, and sentiment tagging across a large corpus of user-generated records. I collaborated on refining annotation guidelines and supported RLHF workflows to enhance model performance and consistency. • Labeled over 10,000 user-generated records for NLP model pipelines. • Contributed to human feedback loops (RLHF) improving model quality. • Reviewed and validated model outputs for AI alignment and safety testing. • Performed data cleaning, de-duplication, and normalization on raw datasets.

As an AI Data Annotation & Training Data Contributor at DXC Technology, I labeled and annotated structured and unstructured text datasets to support machine learning model development. My work included entity recognition, intent classification, and sentiment tagging across a large corpus of user-generated records. I collaborated on refining annotation guidelines and supported RLHF workflows to enhance model performance and consistency. • Labeled over 10,000 user-generated records for NLP model pipelines. • Contributed to human feedback loops (RLHF) improving model quality. • Reviewed and validated model outputs for AI alignment and safety testing. • Performed data cleaning, de-duplication, and normalization on raw datasets.

2022 - 2025

Education

Work History