Alvarado Godinez Kevin Isaac

Key Skills

Software

Appen

CVAT

OneForma

Prodigy

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Text

Top Task Types

Audio Recording

Evaluation Rating

Text Summarization

Translation Localization

Freelancer Overview

I’m a dedicated LLM Evaluator and Data Annotator with hands-on experience creating and reviewing training data for AI models. I’ve labeled over 60,000 examples, including customer-support intents, social media sentiment, and captioning for images. I set up clear guidelines, ran regular check-ins with teammates to keep our work consistent, and used simple scripts to catch any mistakes early on. Alongside labeling, I test and score AI responses from models l, checking for accuracy, clarity, and fairness. I share easy-to-understand feedback that helps improve prompts and model performance. Whether I’m training new annotators, cleaning up data in tools like Label Studio, or helping move approved data into model training pipelines, I focus on steady quality and clear communication.

Labeling Experience

Audio annotation specialist

LabelboxAudioEmotion RecognitionTranslation Localization

In this project, I was responsible for annotating audio data by transcribing speech accurately and identifying speakers and emotions. The scope included labeling various audio clips such as customer calls, medical consultations, and in-car voice commands. Tasks were performed following strict guidelines to ensure high-quality annotations essential for training speech recognition and sentiment analysis models. The project size involved thousands of audio files, and quality assurance processes were adhered to rigorously to maintain accuracy and consistency.

In this project, I was responsible for annotating audio data by transcribing speech accurately and identifying speakers and emotions. The scope included labeling various audio clips such as customer calls, medical consultations, and in-car voice commands. Tasks were performed following strict guidelines to ensure high-quality annotations essential for training speech recognition and sentiment analysis models. The project size involved thousands of audio files, and quality assurance processes were adhered to rigorously to maintain accuracy and consistency.

2023 - 2024

Customer Support Intent Classification

ProdigyTextEntity Ner Classification

Annotated 50,000+ customer chat messages to train a support-bot on common user intents (e.g., “order status,” “refund request,” “technical issue”). Developed detailed guidelines for intent classes and slot tags, ran weekly calibration sessions to maintain ≥ 98% inter-annotator agreement, and implemented a Python QA script to flag inconsistencies and missing entities.

Annotated 50,000+ customer chat messages to train a support-bot on common user intents (e.g., “order status,” “refund request,” “technical issue”). Developed detailed guidelines for intent classes and slot tags, ran weekly calibration sessions to maintain ≥ 98% inter-annotator agreement, and implemented a Python QA script to flag inconsistencies and missing entities.

2022 - 2022

Education

Work History