Karthick V P

Key Skills

Software

AWS SageMaker

Appen

Scale AI

Top Subject Matter

Artificial Intelligence and Machine Learning Data Annotation

Training Data Quality

Annotation quality control

Top Data Types

Image

Text

Document

Top Task Types

Classification

Segmentation

Bounding Box

Point Key Point

Object Detection

Text Generation

Question Answering

Text Summarization

RLHF

Fine Tuning

Red Teaming

Transcription

Evaluation Rating

Computer Programming Coding

Function Calling

Prompt Response Writing SFT

Freelancer Overview

Results-driven AI Data Specialist with hands-on experience in machine learning data operations, LLM evaluation, and large-scale annotation projects. Formerly a Machine Learning Data Associate with Amazon Web Services on the AWS Bedrock team, with additional experience contributing to Amazon annotation initiatives through a vendor engagement. Adept at evaluating large language model outputs across critical quality dimensions — including instruction following, factual correctness, logical consistency, safety compliance, and overall response quality — with a strong focus on delivering high-fidelity training data that drives model improvement. Demonstrated expertise spans prompt generation and response evaluation, hallucination detection, tool usage validation, and multimodal annotation across text and image data. Brings a rigorous, detail-oriented approach to inter-annotator agreement, quality control workflows, and taxonomy-based error classification, ensuring consistent and reproducible labeling standards at scale. Proficient in SQL, Python, and AI evaluation frameworks, with the ability to apply data analysis skills to surface insights, identify annotation patterns, and support continuous quality improvement. Known for analytical thinking, methodical execution, and a commitment to producing training data that meets the highest standards of accuracy and reliability.

Labeling Experience

Machine Learning Data Associate

Aws SagemakerImageClassification

As a Machine Learning Data Associate at Amazon, I enhanced the quality and consistency of data labeling tasks. I implemented advanced training, annotation guidelines, and optimized annotation accuracy across projects. My work involved developing multi-layered quality control systems for improved deliverables. • Led analytics team to improve annotation accuracy from 85% to 95%. • Conducted audits and developed guidelines that reduced error rates by 10%. • Maintained 99% data consistency in quality metrics. • Reduced turnaround time in annotation work by 15% through workflow optimization.

As a Machine Learning Data Associate at Amazon, I enhanced the quality and consistency of data labeling tasks. I implemented advanced training, annotation guidelines, and optimized annotation accuracy across projects. My work involved developing multi-layered quality control systems for improved deliverables. • Led analytics team to improve annotation accuracy from 85% to 95%. • Conducted audits and developed guidelines that reduced error rates by 10%. • Maintained 99% data consistency in quality metrics. • Reduced turnaround time in annotation work by 15% through workflow optimization.

2024 - 2026

Data Analyst I

Aws SagemakerImageClassification

During my role as Data Analyst I at Amazon, I was responsible for annotating large volumes of images and videos with high precision. My contributions significantly improved ground truth data accuracy and supported machine learning model enhancement. I ensured productivity and maintained extremely low error rates throughout the process. • Annotated over 10,000 images and videos at 98% accuracy. • Managed more than 500 files daily with error rates below 1%. • Enhanced data accuracy by 15% and model performance by 20% via precise labeling. • Optimized 1,000+ search queries, improving search relevance by 25%.

During my role as Data Analyst I at Amazon, I was responsible for annotating large volumes of images and videos with high precision. My contributions significantly improved ground truth data accuracy and supported machine learning model enhancement. I ensured productivity and maintained extremely low error rates throughout the process. • Annotated over 10,000 images and videos at 98% accuracy. • Managed more than 500 files daily with error rates below 1%. • Enhanced data accuracy by 15% and model performance by 20% via precise labeling. • Optimized 1,000+ search queries, improving search relevance by 25%.

2023 - 2024

Education

Work History