✳ AI SYSTEMS & CONTEXT ARCHITECT

Human ambition.
System intelligence.

I turn everyday friction into systems that move people forward.

I’m Ryan. I design and build AI-powered products from the first question to the production environment—combining engineering with Industrial-Organizational Psychology.Explore my work

PEOPLECONTEXTCOORDINATED ACTION

INDEPENDENTLY DESIGNED. BUILT END TO END.SCROLL TO EXPLORE

01 / SELECTED WORK

Less friction.
More possibility.

Three systems built around the same conviction: technology should make work easier to do, understand, and trust.

01 — WORKFLOW ORCHESTRATIONLIVE · PRIVATE PRODUCTION

A whole workflow.
Working together.

One support request can become a ticket, a client email, and an internal handoff. Relay coordinates specialized agents, brings the context together, and keeps the important decisions with a person.

RELAY / ORCHESTRATIONILLUSTRATED WORKFLOW · SAMPLE REQUEST

01 / INCOMING CONTEXT

A client needs an update.A support email arrives in Outlook.

Client contextResolved & ready

Connected knowledgeHistory + contacts

02 / SPECIALISTS WORK IN PARALLEL

▤TicketClickUpReady

✉Client updateOutlookReady

⌘Team handoffSlackReady

03 / HUMAN JUDGMENTYour review.
Your decision.

The package pauses for review before external action.

02 — KNOWLEDGE SYSTEMSIN DEVELOPMENT

Helix

Knowledge that keeps
up with the work.

From real application workflows to structured, reviewed documentation. Designed to make complex platforms easier to learn and support.

03 — SYSTEM RESILIENCEWORKING TEST ENVIRONMENT

Sentinel

When an app changes,
learning shouldn’t stop.

A recovery pipeline that detects broken onboarding, diagnoses what changed, and prepares a tested fix for an engineer to review.

CONCEPT → ARCHITECTURE → UX → IMPLEMENTATION

Every project featured here was designed and implemented independently by me, with AI as part of my engineering workflow.

02 / AN EVOLUTION IN HOW I BUILD

Better tools.
Bigger questions.

Relay didn’t begin as a platform. It began with a practical question: how much of the work between the work could I remove?

RELAY / COORDINATED EXECUTION

The breakthrough was
the architecture.

Specialized agents handle different parts of the same request in parallel. Focused context, durable state, and explicit review turn a useful AI workflow into an operational system.

The lesson: design the workflow around the job, then use AI where it adds value.

One request

Ticket agentEmail agentHandoff agent

One coordinated package

Parallel work · focused context · human approval

MEET NAOMI

The next interface is a conversation.

Relay’s voice assistant brings updates and answers into spoken conversation. Today’s architecture provides read access to Relay’s data; broader action across connected platforms is the next ambition.

THE INTERFACE, IN MOTION

A presence
you can explore.

Step inside Naomi’s visual interface. Follow the shift from listening to thinking to speaking.

Violet · Starry night48 FPS · High fidelity · Expressive

Interactive appearance preview

03 / ORIGINAL FRAMEWORK

SYSTEM IMMUNE MEMORY / EVOLVING PRACTICE

A system should remember
what it took to get better.

AI can accelerate a build. Keeping that system understandable is a different challenge.

I developed SIM to preserve the context behind a fix: what broke, why it broke, what worked, and what should catch it next time. Each verified lesson becomes useful context for the next person—or agent.

S I MEXPERIENCE
BECOMES CONTEXT

01 / DETECT

Start with evidence.

Capture the actual failure and its context: the affected component, environment, error, and observable behavior. A clear signal is the beginning of a useful memory.

ILLUSTRATIVE MEMORY RECORD

Signal: A worker task is missing.
Context: A new service was added.
Evidence: Inspect registered tasks and startup logs.

Remember the why.

Keep the root cause, verified resolution, and prevention lesson together.

Recheck what changes.

Old knowledge needs new evidence when dependencies or environments change.

Make context purposeful.

Give each agent the relevant knowledge, with clear ownership and review.

04 / THE FOUNDATION

The human side
has always been there.

Access, motivation, and agency are design concerns. These projects put that belief into practice.

FULL-STACK PRODUCT

Achieve Reentry

A platform for justice-impacted job seekers, bringing job discovery, application tracking, AI-assisted résumés, and administrative insight into one place.

M.S. / INDUSTRIAL-ORGANIZATIONAL PSYCHOLOGY

Leadership & employee outcomes

My capstone proposes a leadership intervention and study design examining motivation and morale—connecting organizational goals with the people doing the work.

RYAN HOPKINS / DESIGNER & BUILDER

05 / THE PERSON BEHIND THE SYSTEMS

I care about
the person using it.

I’m an AI systems and context architect with a master’s degree in Industrial-Organizational Psychology. My work lives at the intersection of human behavior, product design, and engineering.

Working in Customer Success at iT1 Nucleos puts me close to the friction: the repeated task, the missing documentation, the interrupted learner. I design systems around those realities and build them end to end.

My objective is to make useful technology easier to adopt. Relay and Helix are now being developed toward broader use across the company.

BUILDING IN THE CONTEXT OF iT1 Nucleos

DESIGN THE SYSTEM

Workflow architecture
Context engineering
Multi-agent orchestration

BUILD THE PRODUCT

Python · FastAPI · LangGraph
Next.js · React · TypeScript
PostgreSQL · Docker

MAKE IT USABLE

Human-centered UX
Review & approval workflows
Documentation & adoption

OWN THE OUTCOME

Production deployment
Security & access design
QA & operational handoffs

EDUCATION & CONTINUED LEARNING

Human science.
Technical practice.

SOUTHERN NEW HAMPSHIRE UNIVERSITY · 2023–2025M.S. Industrial-Organizational Psychology↗

More credentials ＋

Human ambition.System intelligence.

Less friction.More possibility.

A whole workflow.Working together.