// Full-Stack Software Engineer

William "Zack"
Withers

Building secure, high-throughput systems — and exploring the frontier of AI safety & applied machine learning.

zack@portfolio ~ bash
❯
View Experience Get in Touch

// 01 — About

Engineer at the intersection of data, cloud & AI

I'm a Full-Stack Software Engineer based in Denver, CO with 5 years of experience building secure, high-performance cloud platforms and complex data visualizations. My dual background in Computer Science and Applied Mathematics gives me an edge when problems live at the crossroads of rigorous logic and real-world engineering.

Most recently I've been going deep on AI — not just building with it, but understanding it. I'm actively studying AI safety, transformer architectures, and practical ML engineering, motivated by the belief that the engineers who understand AI systems deeply will be the ones who build them responsibly.

Off the clock I run a home lab — self-hosted servers, network configuration, the kind of tinkering that bleeds naturally into professional work. When I'm not doing that, I'm either on a snowboard or cooking something that probably got too complicated.

🛡️
BlueDot Future of AI
AI Safety Certification · June 2026
RECENTLY COMPLETED
⚙️
BlueDot Technical AI Safety
Application submitted · Targeting ARENA 9.0 thereafter
APPLIED
🧠
How Transformer LLMs Work
DeepLearning.ai · Recently completed
RECENTLY COMPLETED
📐
AI Safety Courses
Multiple providers · Ongoing
ACTIVE STUDY
☁️
AWS Solutions Architect
Amazon Web Services · Associate

// 02 — Experience

Where I've built things

Software Engineer Feb 2023 – Jun 2026
Watchguard Technologies · Seattle, WA (Remote)
  • Led Front-End architecture for secure Cloud Platform initiatives, growing from core contributor to lead engineer.
  • Built an XML parsing pipeline to migrate device network schemas to the cloud, eliminating 90% of manual configuration time across all client onboarding.
  • Directed technical execution of high-impact epics within Agile frameworks, ensuring security compliance and zero-defect deployments.
  • Implemented comprehensive unit and API testing protocols that meaningfully reduced deployment vulnerabilities.
  • Integrated AI coding tools (Cursor, GitHub Copilot) to accelerate delivery on key corporate initiatives.
TypeScript AngularJS XML Pipelines Cloud RESTful APIs Agile
Software Engineer I & Application Developer I Jun 2021 – Feb 2023
Nextworld · Denver, CO
  • Architected a high-performance D3.js visualization engine used across multiple enterprise ERP applications — instantly filtering and drilling into datasets of 1,000+ records.
  • Built end-to-end and unit testing pipelines for every contribution, ensuring zero-defect deliverables for high-stakes Payables applications.
  • Collaborated directly with enterprise clients to translate complex business logic into scalable, user-facing software.
  • Delivered performant features consistently across rapid Agile sprints for enterprise Payables departments.
D3.js JavaScript PostgreSQL Node.js E2E Testing ERP

// 03 — Skills

Technical toolkit

Languages
JavaScript TypeScript Python Java C / C++
Front End
AngularJS D3.js Node.js HTML / CSS
Back End & Cloud
PostgreSQL RESTful APIs XML Pipelines AWS (S3, etc.)
AI & ML
Transformer LLMs AI Safety Cursor / Copilot Prompt Engineering
Practices
Agile / Scrum Unit Testing E2E Testing API Design Secure Deployment
🤖
Currently deep in AI
Beyond building with AI tools, I'm studying the underlying systems: transformer architectures, alignment research, and the engineering challenges of safe AI deployment. Recently completed BlueDot's Future of AI certification and DeepLearning.ai's How Transformer LLMs Work, with multiple AI safety courses ongoing. My goal is to eventually contribute at the intersection of robust systems engineering and responsible AI development.

// 04 — Apps

Things I've built

⛏️
quarry — SQL Workbench for Your Files
Drop CSV, Excel, Parquet, or SQLite files and query them with real SQL — DuckDB compiled to WebAssembly, with a column profiler, notebook, charts, and exports. Nothing ever uploads.
● Live 🔬
GPT-2 Interpretability
GPT-2 small runs entirely in your browser (ONNX Runtime Web), exposing attention patterns, the logit lens, adversarial-suffix comparison, and activation patching. A sibling to the Adversarial Playground.
● Live ⚔️
Adversarial Playground
Live browser demo of how FGSM and PGD adversarial attacks affect a CNN in real time. No server — the model runs entirely client-side.
● Live 🏋️
fit — Training & Recovery
Multi-user WHOOP replacement: recovery & HRV dashboards fed by wearable webhooks, plus an offline-first gym logger with interval timers and e1RM analytics. Serverless AWS (Lambda · DynamoDB · CDK) at ~$0/mo — try the live demo, no account needed.
● Live 🥃
dram — Whiskey Recommender
A whiskey & wine tasting log with a layered recommender — structured similarity, then note-embedding similarity, then an AI-narrated pick. Cognito-gated writes on a serverless AWS backend.
● Live 🧫
petri — Artificial Life Lab
Five self-organizing systems — a million slime-mold agents, particle life, boids, reaction-diffusion, and Lenia — running as WebGPU compute shaders on your GPU, live in the browser.
● Live 🌀
fathom — 3D Fractal Explorer
Raymarched 3D fractals in WebGL2 — six distance-estimated worlds (Mandelbulb, Mandelbox, and friends) with free-flight controls, a dive autopilot, and hi-res image export.
● Live

// More apps shipping soon — building in public

// 05 — Education

Foundation

The University of Alabama
Tuscaloosa, AL · Graduated May 2021 · Honors College · GPA 3.45
B.S. in Computer Science
B.S. in Applied Mathematics