Artificial Intelligence
I build AI systems that reduce repetitive work, support better decisions, and help people move through complex workflows with more clarity and confidence.
Artificial Intelligence
I build AI systems that reduce repetitive work, support better decisions, and help people move through complex workflows with more clarity and confidence.
WHY I’M UNIQUELY SUITED
FOR AGENTIC AI
My nonlinear background gave me something many developers never fully develop: the ability to understand both systems and the people inside them.
Over the years, I worked across web development, design, marketing, production systems, UX, accessibility, and digital strategy. That broad exposure shaped how I think.
I naturally notice patterns, friction points, redundancies, and workflow inefficiencies, while also paying close attention to how people actually behave inside systems.
Agentic AI felt like the first field that naturally combined all of those ways of thinking into one discipline.
HOW MY APPROACH DIFFERS
I don’t just think about whether something works. I think about how it fits into a person’s day, where friction exists, and how systems can become more intuitive, efficient, and reliable over time.
My experience with AI started through creative experimentation. I first used AI heavily for image generation and visual ideation, exploring how models interpreted prompts, styles, and concepts. Over time, that experimentation expanded into content creation and design support for websites and digital projects.
As I continued using AI tools, I started relying on them to help think through difficult design choices, content structure, and communication clarity. I became increasingly interested in how the systems behaved, where they were strong, and where they became unreliable.
When I began using ChatGPT more seriously, I started testing the limits of large language models directly. I noticed hallucinations, inconsistencies, context issues, and behavioral patterns long before I knew the official terminology behind them. Through repeated experimentation, I developed my own understanding of how LLMs process information, where they fail, and how structure and context influence reliability.
Later, when I started formally studying AI and agentic systems, many of the concepts I encountered simply gave names to patterns I had already observed on my own. That process reinforced both my technical understanding and my instinct for systems thinking, experimentation, and workflow design.
HOW I LEARNED AI
WHY AGENTIC AI FELT LIKE THE NATURAL NEXT STEP
It combined everything I naturally think about already: workflows, systems, pattern recognition, human behavior, iteration, edge cases, validation, and structured problem solving.
What drew me in most was not just the models themselves, but the orchestration around them: how tools interact, how workflows are structured, how humans stay involved when needed, and how complex systems can become more predictable and reliable over time.
Select
↓
Plan
↓
Execute
↓
Validate
↓
Test
↓
Commit
This structured loop became the foundation of how I approach AI-assisted development. By breaking work into small, verifiable steps, I’m able to create systems that are more controlled, testable, and reliable instead of relying on vague prompting or guesswork.
I naturally zoom out to understand the larger system before making implementation decisions. I pay close attention to how individual components, workflows, dependencies, and user interactions affect one another over time.
( 01 )
WANDERLUST - THE TRIBUNAL
2035
HOW I THINK THROUGH SYSTEMS
I naturally zoom out to understand the larger system before making implementation decisions. I pay close attention to how individual components, workflows, dependencies, and user interactions affect one another over time.
I naturally zoom out to understand the larger system before making implementation decisions. I pay close attention to how individual components, workflows, dependencies, and user interactions affect one another over time.
I naturally zoom out to understand the larger system before making implementation decisions. I pay close attention to how individual components, workflows, dependencies, and user interactions affect one another over time.
I naturally zoom out to understand the larger system before making implementation decisions. I pay close attention to how individual components, workflows, dependencies, and user interactions affect one another over time.
( 05 )
GARTH THUNDERBOLT - RISKIT MAG
2035
HOW I THINK THROUGH SYSTEMS
AI Newsletter Digest
A semantic deduplication system that transforms overlapping AI newsletters into a cleaner digest of distinct stories using embeddings, LLM-based clustering, and structured filtering workflows.
[ View Project ]
Deterministic Agent Workflow
A structured AI development system focused on validation, testing, and repeatable execution.
[ View Project ]
AI Portfolio Chatbot
An AI assistant that answers questions about my background while demonstrating structured, reliable agent behavior.
[ View Project ]
AI PROJECTS BUILT FROM THIS THINKING
-
I look for hidden friction, repeated failure points, and operational inefficiencies that emerge across workflows, not just isolated technical problems.
-
I think carefully about how systems behave under real-world conditions, including edge cases, incomplete data, inconsistent inputs, changing requirements, and human unpredictability.
-
I prioritize structure, validation, and observability because complex systems become difficult to improve when behavior cannot be clearly traced, tested, or verified.
-
I pay close attention to details because small inconsistencies often reveal larger architectural or workflow problems underneath the surface.
-
When evaluating AI tools, frameworks, or workflows, I focus less on hype and more on tradeoffs, reliability, integration complexity, and whether the tool genuinely improves the overall system.
• • I enjoy solving problems that require balancing technical implementation, operational constraints, and human usability at the same time.
I naturally zoom out to understand the larger system before making implementation decisions. I pay close attention to how individual components, workflows, dependencies, and user interactions affect one another over time.
( 05 )
GARTH THUNDERBOLT - RISKIT MAG
2035
I naturally zoom out to understand the larger system before making implementation decisions. I pay close attention to how individual components, workflows, dependencies, and user interactions affect one another over time.
( 05 )
GARTH THUNDERBOLT - RISKIT MAG
2035