# Skills

I am strongest where product judgment, messy systems, and hands-on shipping
have to live in the same person.

This file now tracks the left-column skills inventory from my resume more
directly.

## AI/ML

Machine Learning, Deep Learning, NLP, Generative AI, LLM Applications, AI
Agents, Prompt Engineering, Transformers, OpenAI API, Agentic Workflows, Model
Evaluation, XGBoost, Scikit-learn, TensorFlow, PyTorch, Model Deployment,
Feature Engineering.

## Python

FastAPI, Flask, Django, Pandas, NumPy, REST APIs, SQLAlchemy, Asyncio, Backend
Development, Automation, Data Science, Pytest, Unit Testing, Celery, Web
Scraping, OOP, Scripting.

## Frontend

Next.js, Native iOS (Swift), Native Android (Kotlin), Flutter, React,
JavaScript, TypeScript, HTML, CSS, Tailwind CSS, Bootstrap, Responsive Design,
Accessibility.

## Data Engineering

Data Pipelines, ETL, ELT, SQL, PostgreSQL, MySQL, SQLite, Snowflake, Data
Modeling, Data Warehousing, Analytics Engineering, Workflow Automation, Batch
Processing, Scalable Data Systems.

## CLI/TUI Tooling

Python curses, Rich, Textual, Keyboard Event Handling, Interactive Console,
Shell Scripting.

## System Design / Delivery

System Design, Scalable Systems, Architecture Design, API Design, Performance
Optimization, Distributed Systems, Project Management, Technical Leadership,
Agile Methodologies, Risk Management, Delivery Management.

## Unix

Linux, Shell Scripting, Bash, CLI, TUI, Git, SSH, Cron, Sed, Awk, Grep, Vim,
System Administration.

## Cloud

AWS, S3, Docker, Kubernetes, CI/CD, Terraform, Serverless, AWS Lambda, Google
Cloud, Azure, Microservices, Infrastructure as Code, Scalability.

## How Those Skills Actually Show Up

- turning rough operator pain into usable product surfaces before an
  organization invents more ceremony around the problem
- reducing legacy sprawl into something smaller, clearer, and easier to reason
  about
- finding the real source of truth inside an undocumented system and rebuilding
  around that instead of around mythology
- choosing the right container for the work, whether that is a web surface, a
  CLI/TUI, or a terminal-native runtime like `erza`
- moving from concept to shipped tool fast without treating planning as a
  religion

## Public Repo Evidence

- `erza` -> repo: https://github.com/ryangerardwilson/erza | docs:
  https://erza.ryangerardwilson.com/ | terminal-native UI language and runtime
  for docs, tools, local apps, and remote surfaces
- `xyz` -> repo: https://github.com/ryangerardwilson/xyz | outcome-first task
  tracking with JTBD structure and North Star scoring
- `vixl` -> repo: https://github.com/ryangerardwilson/vixl | Pandas- and
  NumPy-backed spreadsheet work in a terminal UI
- `gvim` -> repo: https://github.com/ryangerardwilson/gvim | GTK4 block editor
  with Vim as the editing core, Python renders, and static publishing
- `o` -> repo: https://github.com/ryangerardwilson/o | keyboard-first file
  navigation, picker/save flows, reveal mode, and handler-based opening
- `rt` -> repo: https://github.com/ryangerardwilson/rt | terminal tutor for
  drills, pattern hardening, editing, and search
- `Personalized Agentic Tooling Stack` -> components:
  https://github.com/ryangerardwilson/slack,
  https://github.com/ryangerardwilson/gmail,
  https://github.com/ryangerardwilson/gdrive,
  https://github.com/ryangerardwilson/gcal,
  https://github.com/ryangerardwilson/replyguy,
  https://github.com/ryangerardwilson/x,
  https://github.com/ryangerardwilson/linkedin,
  https://github.com/ryangerardwilson/blog | shell-first communication,
  backup, calendar, posting, and agent-assisted workflow glue

## Working Style

- I am comfortable across product, engineering, data, and copy.
- I usually prefer one sharp tool over five vague platforms.
- I do my best work when the brief is underspecified but the constraint is real.
- I distrust impressive architecture that does not improve the lived workflow.