Tomas Mejia

Key Skills

Software

Internal/Proprietary Tooling

Top Subject Matter

Healthcare – Medical Terminology & Clinical Conversations

Artificial Intelligence – LLM Evaluation & Prompt Quality

Software & Technology – App Development & Data Pipelines

Top Data Types

Image

Text

Document

Top Task Types

Fine Tuning

Classification

Segmentation

Freelancer Overview

I have hands-on experience designing and implementing AI data pipelines that required the same core skills central to data labeling and AI training work: precision, structured thinking, and a deep understanding of how data quality directly impacts model performance. In my PlantCare AI project, I built a full RAG pipeline from scratch — handling data chunking, embedding generation with sentence-transformers, cosine-similarity retrieval, and vector storage — which gave me direct exposure to how training data must be structured and curated for LLMs to produce reliable outputs. My work with multi-agent LLM systems using LangChain and Gemini API also required rigorous prompt engineering and continuous evaluation of model outputs for accuracy, relevance, and consistency — tasks that closely mirror real-world AI training data workflows. What sets me apart is the combination of technical AI/ML expertise with native-level bilingual fluency in Spanish and English (C1 certified). My two years of professional experience as a Medical Interpreter trained me to catch subtle linguistic errors, maintain strict quality standards, and operate with precision in high-stakes, zero-tolerance-for-mistakes environments — qualities that translate directly into high-quality data annotation and model evaluation. I am already deeply familiar with AI development tools such as Claude, Cursor, and GitHub Copilot, and I understand how human feedback shapes model behavior, making me an informed and effective contributor to any AI training data pipeline.

Labeling Experience

Education

Work History