● Quick view →
.pdf.docx.doc.rtf.txt

Data & Automation Infrastructure Engineer

Justin Parsons

Automation Engineer | Data Infrastructure & Pipeline Mechanics

📍 Cocoa, FL↗ justinsamuelparsons.com

Data & Automation Infrastructure Engineer with 10+ years running industrial packaging and controls systems, now applying the same discipline to data. I build containerized PostgreSQL environments and Python-driven pipelines that automate schema migrations, batch ingestion, and analytic cost reporting, treating every pipeline like a production line: instrumented, repeatable, and built to run without babysitting.

01

Skill Matrix

Click any skill to light up the projects where I used it. Numbers show how many.

Languages

Databases & Pipelines

Tools & Infrastructure

Web & Application

AI & Machine Learning

Industrial Controls & Electrical

Mechanical & Design

Process & Quality

02

Highlight of the Year

Project:

Data-Automation-Pipeline-V1

End-to-end data pipeline that uses Python connection drivers (psycopg2) to reach through the network layers of a local, Docker-hosted PostgreSQL instance and fully automate the database lifecycle.

Pythonpsycopg2PostgreSQLDockerWSL2SQL

PostgreSQL runs inside a Docker container with mapped ports and persistent volumes, so the whole environment can be torn down and rebuilt identically on any machine.

Python connects across the container's network boundary through psycopg2, handling host/port mapping, credentials, and transactional cursors from WSL into the container.

Scripts drop stale objects and apply schema migrations in order, creating tables, keys, and constraints so the relational model is rebuilt idempotently on every run.

Source records are loaded in batches inside transactions, so a failed batch rolls back cleanly instead of leaving partial data behind.

Aggregation queries calculate cost metrics directly in the database and return reporting-ready results to Python.

  • ✓ One command rebuilds the database from zero to analytics
  • ✓ Repeatable, environment-independent runs via Docker
  • ✓ Transaction-safe ingestion with rollback on failure
03

Additional Builds

CAD / Web / Pricing

Factory Spec Builder

Browser tool that parses uploaded STEP/CAD files with WebAssembly, renders them in interactive 3D, measures part dimensions, and runs a cost-matrix engine for material, process, and assembly pricing through to checkout.

React Three FiberWebAssemblyTypeScriptStripe

Automation / Self-hosting

Self-Hosted Promo Pipeline

Automated content pipeline running from a home server: Python and SQLite manage a weighted rotation schedule, delivered securely through a Cloudflare Tunnel with origin validation.

PythonSQLiteCloudflare TunnelLinux

Local AI / Orchestration

Local AI Video Studio

Fully local generative workflow: a Gemma model on Ollama develops prompts, a Streamlit controller directs the pipeline, ComfyUI renders video, and results are cataloged in a self-hosted media library.

OllamaGemmaStreamlitComfyUI

Generative / QA process

AI Artwork Workflow

Repeatable AI image pipeline with written briefs, brand-consistency rules, review checkpoints, and optimized delivery formats, documented step by step.

AI Image GenerationCompositingWebP Delivery
04

Build Log: Progress with Lovable

Everything I've shipped since I started building with Lovable, in order, with the skills each one put to work.

  • Personal Landing Page

    First Lovable build: a one-page site introducing my engineering, AI, and creative disciplines, published live.

  • Business Portfolio & Storefront

    Grew into a multi-page business site: portfolio galleries, two resumes, a quote and custom-request system with email, a merch store with Stripe checkout, and a self-hosted media feed.

  • Business Card Generator

    Tool that lays out printable business cards (front and back, horizontal and vertical) from a profile of skills and contact details.

  • Streaming Hub

    Central page for streaming content with integrated social and channel links, published live.

  • Signal Identity Creator

    Brand-system tool that defines color tokens, typography, and identity pages, with AI-generated artwork held to written art-direction rules.

  • Factory Spec Builder (inside the business site)

    3D CAD upload and quoting tool: parses STEP files in the browser, measures parts, and prices material, process, and assembly.

  • Data & Automation Resume (this page)

    Recruiter-ready portfolio with an interactive skill matrix and one-click PDF, Word, RTF, and text resume exports.

05

Experience

Data-focused roles, 2021 onward.

  1. 2025 - Present

    Data & Automation Engineer (Independent)

    • ▸ Built Data-Automation-Pipeline-V1: containerized PostgreSQL with Python-driven migrations, batch loads, and analytic SQL.
    • ▸ Built a self-hosted promo pipeline on a home server with Python, SQLite, weighted rotation, and Cloudflare Tunnel delivery.
    • ▸ Run Linux, Docker, and WSL/hypervisor environments with Git-based workflows.
  2. 2021 - 2026

    Engineering Technician, Hydro Precision Tubing (Rockledge, FL)

    • ▸ Installed new equipment end to end: power, line controls, pneumatics, mechanical assembly, and fabrication.
    • ▸ Integrated measuring equipment with analog/digital I/O for temperature, flow, and pressure data.
    • ▸ Used software data readings to link process and quality problems, driving fixes through to stable full-time production.
    • ▸ Diagnosed electrical faults and edited PLC controls logic.
06

From the Production Line to the Pipeline

I learned automation on the factory floor before I wrote a line of Python. For more than a decade I kept high-speed packaging lines running: tuning film-web tension, dialing in induction sealing, fabricating parts that didn't exist, and editing Allen-Bradley PLC ladder logic so machines did what the process needed.

That work taught me how systems actually fail: at the handoffs, under load, at 3 a.m. Data pipelines fail the same way. So I rebuilt my toolkit around Linux, Docker, PostgreSQL, and Python, and I design pipelines the way I'd commission a machine: isolate it, instrument it, make every step repeatable, and make recovery boring.

I'm looking for a founding or early team where infrastructure has to be dependable from day one, and where someone who has run physical production systems can bring that rigor to data.

Justin Parsons · Cocoa, FL · justinsamuelparsons.com
Edit with