Building AI Systems
for Real Business
Operations.
Exploring how AI agents, automation, and human workflows can evolve together — documented by a business operator with 20 years in the field.
Meet Adina
She looks a little familiar.
Adina is my interactive AI avatar and your guide to Cultivated Cure. Ask her where to start with AI, what I've been experimenting with, how I use different AI tools, or just let her show you around.
She's AI, so she can occasionally get things wrong. Think of her as a conversational guide to my experiences and experiments — not an oracle.
A Note on Adina
Adina is an AI-generated guide based on Teresa's public content and experiences. AI can make mistakes. Please don't share confidential or sensitive information. Her responses are educational and are not professional advice or a guarantee of results.
Battle of the Bots
Two autonomous AI agents. Opposite strategies. Real money on Coinbase. Round 1 went the distance — decided by which bot bled less through a brutal downturn. Round 2 opens Sept 1 with a new format.
Operations Workflow Architecture
Sanitized conceptual models for delivery readiness, escalation paths, and AI-assisted operational forecasting. Exploring what human-in-the-loop systems look like at scale.
Project Sagebrush
AI-assisted coordination for a ranch house restoration — schedules, budgets, and next steps proposed by GPT-4, never executed without a human decision. Built on Airtable and Notion.
Gmail Organization Agent
Personal productivity agent for email triage, categorization, and automated response workflows with human oversight and override controls built in.
AI Presentation Systems
Building AI-assisted presentations, avatars, and interactive learning experiences. Exploring how AI changes education and communication at scale.
Synthetic Media Lab
Exploring AI-generated images, videos, and digital media for storytelling and education. Building a personal toolkit for synthetic content creation.
Family Trip Planners
How a "quick brochure" request turned into two full interactive trip-planning apps. Two live, working planners embedded — built through pure feature creep.
If AI dramatically increases knowledge production, what becomes the new organizational bottleneck?
- Hypothesis
- AI accelerates systems. It does not automatically repair them.
- What Challenged It
- A counterexample keeps nagging at me: sometimes speeding up one broken step forces a team to finally fix the handoff around it, because the mess becomes too visible to ignore. Speed can also be the thing that exposes the repair, not just the thing that skips it.
- Exploring Next
- Whether 'decision latency' can be measured the same way cycle time is measured on a shop floor.
The Framework
The AI Throughput Lens
A five-part framework, open research questions, and the journey from automation to this research.
In Development
The Throughput Mindset
"There is a particular kind of exhaustion that sleep doesn't fix." — Chapter One: The Wrong Problem.
Teresa Lechuga
Bridging Operations
and AI to Build
Smarter Systems.
I'm a business operator and systems thinker with 20+ years in commercial operations — most of it spent inside a multi-region territory, owning P&L, marketing, and technology initiatives for a large organization. I didn't come into AI from a technical background. I came into it the way most of the interesting turns in my career have happened: something amazed me, and I immediately went to go break it on a real problem.
I hold an MBA from Grand Canyon University and a Master's in Legal Studies from Arizona State University, and I lean on that combination — business strategy, legal reasoning, and two decades of operational pattern-matching — every time I sit down to build something. This site documents practical experimentation, not theory: the AI Throughput research, Battle of the Bots, Project Sagebrush, and whatever's next.
- +Operations Leader
- +AI Experimenter
- +Workflow Architect
- +Problem Solver
Outside the Systems
Outside the Systems
Experience
20+ Years Operations
Education
MBA (GCU) · MLS (ASU)
Focus
AI Systems · Workflow Architecture
Based In
Northern California
Why Every Improvement Creates Another Bottleneck
This is the note that started the rest of them. I kept closing out automation projects that worked exactly as designed, and the organization didn't get faster in any way I could point to.
Read Full Note →AI Is My Jonah
I said it as a joke in a chat window before I meant it. Somewhere in a few hundred back-and-forth exchanges, I noticed AI doing the same thing to me that Jonah did to Alex Rogo.
Read Full Note →Throughput Is Not Speed
If everyone works faster, does more value actually reach the customer? Not automatically — and that gap is the whole reason this research exists.
Read Full Note →Building Trust in AI Systems One Decision at a Time
Accountability isn't a feature you bolt on later. It has to be designed from the start — visible, intentional, and human-centered.
Read Full Note →The future isn't human or AI.
It's both.
We are not being replaced.
We are becoming augmented.
AI won't replace you. But an augmented you might.
Be the Operator
Ask better questions. Challenge the output. You are still the decision maker.
Validate Everything
AI is confident. Confidence is not the same as truth.
Embrace. Share. Grow.
Don't hide the tools that help you build. Share what works.
Stay Human.
Technology changes. Humanity remains. Keep showing up.
Interested in
Practical Conversations
About AI Systems?
I'm open to conversations about AI in operational environments, workflow design, practical experimentation, speaking, and meaningful collaboration.
Also on
LinkedIn → Teresa Lechuga