Skip to content

Who I am

I’m Jeremy Dill, an AI Creative Systems Specialist based in Maryland. I’ve spent the last 18 years working across design, web, motion, 3D, and interactive production. I came into AI through making things, not through theory.

What I do

I build creative production systems that help teams control what happens before, during, and after generation. My work combines strategy, interface design, automation, creative direction, and QA to turn unclear inputs into work people can review, trust, and use.

My foundation

Before Aiuci, I worked at the American Geophysical Union across graphic design, game development, and motion graphics. I later built an 800-artist public profile system for LED Baltimore. That experience in real creative production became the foundation for the AI systems I build today.

AI Creative Decision Assurance

I make AI-assisted creative work safer to trust, easier to inspect, and less painful to run.

Make Output Earn Trust

Creative chaos gets expensive fast.

Creative work slows down after the output looks finished. Every weak claim creates rework. Every lost decision creates delay. Every hidden blocker creates another meeting. I build production systems that pressure-test output, preserve decisions, connect owners, and turn repeat creative work into workflows a team can actually run.

Stress tests that kill weak output.

Work gets pressured before it ships, so bad logic, missing proof, and false confidence do not become expensive rework.

Output to proof

360-degree audit logic.

I force AI work through opposing audit lenses until blind spots, weak assumptions, and false confidence surface before shipping.

Blind spots to proof

Control & accountability.

Briefs, assets, reviews, blockers, owners, and handoffs connect into one visible system, so the work has an owner, a status, and a reason it moved.

Chaos to system

Safer equals more.

When repeat creative work becomes inspectable, reusable, and governed, teams can run more of it with less fear.

Repeat to workflow

Companies do not avoid AI because they hate speed. They avoid it because no one wants to be accountable when the system does something stupid at scale. Safer equals more.

Feweragent babysitting hours
Fasterasset-to-action delivery
Cleanervalidated deliverables
SaferAI-assisted production
“ Less agentic overhead because the system shows what survived, what was decided, who owns it, and why it moved. Make It Flow I turn scattered creative output, approvals, and follow-ups into an accountable operating system a team can actually run.
The operating layer
□Stress ○Decisions ✦Systems △Status #Ownership ↻Flow

Skills + Systems

Roles
  • Creative Automation Specialist
  • AI Creative Systems Builder
  • Agentic Workflow Systems Designer
  • Creative Technology Specialist
  • AI Content Systems Specialist
  • AI Video Workflow Builder
  • Front-End Creative Developer
Capabilities
  • Creative automation architecture
  • Agent workflow spec design
  • Prompt and context systems
  • QA and drift control
  • Review gates and approval flows
  • Controlled image and video variation
  • Identity preservation
  • Dashboard and interface design
  • React / Next.js workflow interfaces
  • Structured AI outputs
  • Production handoff systems
  • Audit trail design
Web + Front-End

I build the interface layer around the workflow: responsive sections, control-room views, dashboards, component-based layouts, JavaScript behavior, React / Next.js experiences, and AI-assisted production code that still depends on structure, QA, and deployment judgment.

WordPress
JavaScript
React
Next.js
HTML / CSS
Responsive UI
Component Systems
Interactive Front-End
Agents + Workflow Logic

I design agentic systems as job engines: trigger-based workflows, context windows, tool permissions, approval gates, escalation rules, eval sets, control-room interfaces, and repeatable product wrappers that make AI labor visible, testable, and safe to use.

Automated Workflow Software
AI Intake Systems
Draft + Approve Flows
Review Gate Systems
Approval Queues
Human Review Gates
Escalation Paths
QA Scorecards
Agent Test Sets
Audit Logs
Dashboard Interfaces
React Dashboards
Custom Agent Skills
Custom Agent Plugins
AI Tool Workflows
Structured AI Outputs
API Handoffs
AI Coding + Agents
Codex
Claude Code
Google Antigravity
Hermes
NemoClaw
Gemma
Music + Sound
Ableton Live

More output is not the advantage. Controlled output is.

The AI Creative Control Problem

AI creative output is exploding. Teams cannot control it. They cannot repeat it. They cannot QA it. They cannot safely hand it off.

Interactive role-to-proof map

Choose a role. See the proof.

Select one or more roles in the top row. The lines reveal the projects below that prove each capability.

1 — Choose a roleSelect more than one to combine them
Selected roles / 01

What changes for the team

Controls exploratory AI output until it becomes consistent and reproducible.

2 — Inspect the projects that prove it
PRJ—01

Tek Sports Insights

Product architecture, naming, seven homepage directions, and a branded sports asset system.

Inspect proof
SYS—02

Scope Logic

Variation architecture, prompt controls, QA gates, rejection logic, and approved output families.

Inspect proof
EXP—03

Emotion + Motion

A repeatable translation layer from emotional intent to motion behavior and production rules.

Inspect proof
LAB—04

Playcio

Concept development, visual systems, interactive prototyping, and production-ready experiments.

Inspect proof
ARC—05

Design, Motion & 3D

Direct evidence of taste, execution range, finishing standards, and hands-on craft.

Inspect proof
WEB—06

Web + Production

Customer-facing interfaces and campaign systems carried from idea through delivery.

Inspect proof
AI Execution Control
360° audit before agents run
Scope Logic · system proof

Bad direction gets caught before production starts.

AI is moving faster than people can inspect what it produces. I built Scope Logic to use AI as an auditing system before it becomes a production engine.

Four-quadrant reasoning creates a complete view of the problem, while an equation layer converts linguistic ambiguity into deterministic decision data before anything runs.

Unclear Audited Ready
Loose Brief
Messy Notes
AI Prompt
Client Request
360°
Four-Quadrant Audit Builds the full view Forces the model to inspect the problem from four angles before it acts.
Objective Locked
Ambiguity Resolved
Ready to Run
Live Problem mapped
Reasoning quadrants 4
Control gates 5
Source of truth 1
360° Problem view before execution
A→D Ambiguity converted into deterministic data
Any Domain, workflow, or decision type
Use intelligence to audit the work before automation accelerates it.