About Me

Analytics Engineer with experience building data platforms at scale in manufacturing environments. At Hankook Tire I built ePMT, an internal MES analytics platform with a PostgreSQL warehouse, dbt transformation layer, and automated ELT pipelines, reducing manual reporting time by 90% and giving plant leadership live visibility into production metrics.

Background in Industrial Engineering and Information Systems. I understand warehouse design, transformation logic, and the operational context that makes data actually useful to the people reading it. Published researcher in applied deep learning, ResNet50 CNN, 97.9% accuracy, deployed on GCP.

On the side I work with small and medium businesses to replace manual operations with systems that are clean, reliable, and tailored to how they work.

Impact

  • 90% reduction in manual reporting time PostgreSQL · dbt · Python · Airflow · MES
  • 60 machines covered by scheduling pipeline Python · MongoDB
  • 7% cycle time variance reduction PostgreSQL · dbt · Python · MES · IIoT
  • 97.9% test accuracy on published CNN model Python · ResNet50 · GCP

What I Build

  • ELT Pipelines & Warehouses

    ELT pipelines from source to warehouse, with a dbt transformation layer on top. Star schema and medallion modeling. Built and tested in production against MES and IIoT data at Hankook.

    PythonPostgreSQLdbtAirflowSQL

  • Analytics Platforms

    Operational dashboards and reporting systems that leadership and operations teams actually use. OEE, cycle time, KPI tracking. Replaces spreadsheets with data that drives decisions.

    dbtGrafanaPostgreSQLPython

  • Operational Web Apps

    Internal tools and custom applications that replace disconnected Excel files and manual workflows. Admin dashboards, digital forms, data entry and query interfaces built to fit exact operations.

    Next.jsReactPostgreSQLPython

Certifications

Resume

Experience

  1. Hankook Tire Manufacturing Tennessee

    Production Management Engineer I · Analytics & Data Engineering

    Mar 2025 — Present · Clarksville, TN

    Reduced manual reporting time by 90% by engineering ePMT, an internal MES analytics platform with automated ELT pipelines and a dbt transformation layer, giving plant leadership live visibility into OEE, cycle time, and inventory.

    Automated plant-wide scheduling audits by architecting a Python and MongoDB pipeline that reconstructs historical production states across 60 machines simultaneously, replacing manual operator snapshots with a queryable historical interface and reducing audit prep from days to seconds.

    Architecting an operational platform for the Mold & Bladder department using PostgreSQL, dbt, and Airflow, designing data models, digital forms, and automated workflows to replace manual mold tracking and fragmented Excel reporting.

    Reduced cycle time variance by 7% by developing an ELT pipeline across MES and IIoT machine data into PostgreSQL, computing sub-process metrics with dbt across individual production stages, and presenting anomaly findings to drive operator retraining.

    PostgreSQLdbtPythonAirflowMongoDBELTMESSAP

Education

  1. University of Arizona

    B.S. Industrial Engineering — GPA 3.73 · Magna Cum Laude

    Mar 2021 — Dec 2024 · Tucson, AZ

    Relevant coursework: Production Systems Analysis, Simulation Modelling, Engineering Decision Making, Project Management.

  2. Universidad Peruana de Ciencias Aplicadas

    B.S. Information Systems Engineering

    Mar 2021 — Dec 2025

    Relevant coursework: Web Applications Architecture, Database Design, Networks & Communications.

Publication

  1. Cogent Engineering, Taylor & Francis

    Deep learning-based quality control system for meat products

    Nov 2025 · DOI: 10.1080/23311916.2025.2579705

    Paredes-Molina, E., Flores-Zuazo, R., & Carrera-Salas, J. ResNet50 CNN achieving 97.9% test accuracy, deployed on GCP.

    Deep LearningResNet50GCPComputer Vision

Technical Skills

  • Data & Warehousing
    SQLPostgreSQLOLAPData ModelingdbtMedallion ArchitectureELT
  • Infrastructure & Orchestration
    Apache KafkaApache AirflowDockerGCPBigQueryAWS
  • Development
    PythonNext.jsReactFull-stack Web AppsMES Integration
  • Analytics & BI
    GrafanaPower BIOEE AnalysisCapacity ModelingRoot Cause Analysis
  • Domain
    Manufacturing OperationsIIoTMES PlatformsSAPBPMN
  • Certifications (In Progress)
    AWS Certified Data Engineer Associatedbt Certified Developer

Work

Production systems built at Hankook Tire Manufacturing Tennessee, a top-10 global tire manufacturer with $6B+ annual revenue. These are not portfolio projects. They run in production and are used daily by plant leadership and engineering teams. No GitHub links — proprietary.

Production Systems

  1. ePMT — MES Analytics Platform

    Mar 2025 — Present · Hankook Tire, Clarksville, TN

    Internal analytics platform connecting the plant's MES (Manufacturing Execution System) to a self-built PostgreSQL warehouse with a dbt transformation layer and automated ELT pipelines. Reduced manual reporting time by 90%. Gives plant leadership live visibility into OEE, cycle time, and inventory. Solo owner of all data infrastructure. IT not involved.

    PostgreSQLdbtPythonELTMESOEE

  2. Scheduling Intelligence Pipeline

    2025 · Hankook Tire, Clarksville, TN

    Python and MongoDB pipeline that reconstructs historical production states across 60 machines simultaneously. Replaced manual operator snapshots sent by email with a queryable historical interface. Reduced scheduling audit prep from days to seconds. Foundation for future ML-based scheduling optimization.

    PythonMongoDBHistorical ModelingScheduling

  3. Mold & Bladder Operational Platform

    In Progress · Hankook Tire, Clarksville, TN

    Operational platform for the Mold & Bladder department using PostgreSQL, dbt, and Airflow. Replacing manual mold tracking (engineers lose molds for days or weeks) and fragmented Excel reporting across engineering and operations teams with centralized data models, digital forms, and automated workflows.

    PostgreSQLdbtAirflowDigital FormsWorkflow Automation

  4. Cycle Time Variance Analysis

    2025 · Hankook Tire, Clarksville, TN

    ELT pipeline pulling from MES and IIoT machine data into PostgreSQL. dbt models compute sub-process metrics across individual production stages. Anomaly findings presented to the manufacturing team, driving operator retraining that reduced cycle time variance by 7%.

    PostgreSQLdbtIIoTELTRoot Cause Analysis

Portfolio

Services

On the side I work with small and mid-size businesses to replace manual chaos with systems that are clean, reliable, and built for how they actually work. Currently working with two clients. Taking on new projects selectively.

  • Digital Transformation

    I come in, learn how your business actually works, and identify where time and money are being lost to manual processes, disconnected tools, and information that lives nowhere useful. Then I build a roadmap and execute it. The result is a business that runs on reliable data and systems that work without constant human intervention.

    Process audit Systems integration Automation Staff handoff

  • Process Improvement

    If your team does the same things over and over but nobody has ever sat down and mapped out exactly how, there is almost certainly a faster, cheaper, less error-prone way to do it. I document your current processes using BPMN, find the waste, redesign the flow, and implement the changes. You keep what works. You stop doing what does not.

    BPMN mapping Workflow redesign Error reduction SOPs

  • Custom Software and Web Apps

    Sometimes the right tool does not exist yet, or the ones that do are built for companies ten times your size and cost accordingly. I design and build full stack web applications tailored to your exact workflows. Whether you need an internal tool your team uses every day, or a customer-facing app that represents your business online, it gets built for you, not for a generic use case.

    Web apps Admin dashboards Internal tools Full stack

Currently working with two clients. Taking on new projects selectively.

Get in touch

Contact

Send a Message

0 / 1000