REF: DATA ENGINEERING · DEVOPS · AI SYSTEMS LOC: MÉRIDA, MX · UTC-6

ELIO EDUARDO UCÁN ZAPATA BUILDING DATA PIPELINES & AI

Data Engineering & AI student at UPY specializing in DevOps at Holberton School, currently conducting AI Research at The Palace Company. Focused on architecting reproducible ETL workflows, automated cloud infrastructure on AWS, and production-ready AI agent APIs.

DOWNLOAD CV [PDF] GITHUB LINKEDIN

05 Projects Shipped Production & Open Source

06 Certifications AWS & DataCamp Verified

AWS Certified Engineer Data & Cloud Foundations

C1 English · iTEP 4.7 Professional Working Fluency

THESIS // ABOUT

I build the foundational data infrastructure layer: ingestion pipelines, cloud environments, and the APIs that bind them into reliable systems. Based in Mérida, México, studying data engineering full-time while designing and shipping production-grade architectures.

My focus is deterministic pipelines, strict schema enforcement, and cost-effective cloud deployments with Infrastructure as Code. Every project reflects rigorous engineering craft, zero unnecessary bloat, and complete reproducibility.

NOW // ACTIVE FOCUS

[WHAT I'M RESEARCHING & BUILDING · FALL 2026]

[01] RESEARCH

RAG & Reranking Optimization

Investigating reranking algorithms and context-window token economics to maximize retrieval precision in agentic AI architectures.

[02] INFRA

Self-Hosted Server Automation

Running continuous workflows on a headless Debian server with n8n and containerized AI tasks under strict 4 GB RAM constraints.

[03] CLOUD

Cloud Data Engineering on AWS

Deepening hands-on data architectures on AWS (S3, Glue, Athena, Redshift) driven by reproducible Infrastructure as Code and automation.

[04] WRITING

Internship Technical Thesis

Authoring a comprehensive 40-page research and systems report in LaTeX detailing production AI architectures and pipeline benchmarks.

FEATURED PROJECTS

[05 ENTRIES INDEXED · ALL REPRODUCIBLE]

[01]

DC Operations Dashboard

PRODUCTION

Python Streamlit Plotly Docker Terraform AWS

Multi-page operational dashboard for a 100 MW data center — uptime SLA (99.982%), PUE trends, and TIA-942 / ISO 27001 compliance across 5 views. Deployed on AWS with Terraform IaC and GitHub Actions CI/CD.

99.982% SLA · Terraform IaC Automated

CODE & DEMO 2026

[02]

ETL Pipeline — Airflow + Docker + PostgreSQL

DATA PIPELINE

Python Airflow Docker PostgreSQL pandas

Production-grade daily ingestion pipeline consuming the CoinGecko REST API with under 1% failure rate via Airflow retry logic. Zero-config deployment with Docker Compose.

<1% Failure Rate · Daily Automated Ingestion

CODE 2025–26

[03]

Infrastructure as Code — Terraform on AWS

CLOUD IAC

Terraform AWS EC2 S3 IAM

Reproducible AWS environments (EC2 + S3) using modular Terraform with remote state. Reduced provisioning time by ~90% — from hours to under 10 minutes.

-90% Provisioning Time · Modular Remote State

CODE 2025

[04]

AI Agent — Anthropic API + FastAPI

AI SYSTEMS

Python FastAPI Docker Anthropic SDK AWS EC2

Conversational agent with live tool use (CSV reader + PostgreSQL queries) and persistent memory, served as a production REST API. Deployed to AWS EC2 with Docker Compose.

Live Tool Use · Persistent PostgreSQL Memory

CODE 2025–26

[05]

better-readmes

OPEN SOURCE

Anthropic SDK Claude Code npm/npx TypeScript

Published open-source Claude Code skill that auto-generates READMEs and fills GitHub repository metadata. Distributed via npm/npx — usable by any Claude Code user globally.

Global npm/npx Distribution · Claude Code Skill

CODE 2026

PROFESSIONAL EXPERIENCE

[02 ROLES RECORDED · APPLIED ENGINEERING]

[01]

AI Research Intern

The Palace Company ACTIVE ROLE

AI Partner Research Technical Negotiation RAG & Embeddings Python Anthropic API Docker

Leading technical research and comprehensive benchmark evaluations of diverse AI partners and vendors to drive multi-area enterprise AI integrations.
Liaising between cross-functional business stakeholders and technology providers, participating directly in technical negotiations and solution selection.
Researching and prototyping Retrieval-Augmented Generation (RAG) architectures with semantic embeddings, reranking, and token-cost optimization.
Designing automated AI workflows and internal APIs to accelerate intelligence across diverse operational departments.

Sep 2026 – Dec 2026 Mérida, México · On-site

[02]

DevOps Engineering Fellow

Holberton School Mérida FELLOWSHIP

C Bash Linux Docker Git CI/CD AWS

Mastered low-level systems programming fundamentals in C, POSIX system calls, memory management, and Unix system administration with Bash.
Engineered automated CI/CD pipelines, containerized microservices with Docker, and provisioned modular cloud infrastructure on AWS.
Awarded a 50% merit scholarship after securing 3rd place in the Holberton Programming Olympics 2025.

Jan 2025 – Present Mérida, México