01 Where could AI be useful?
AI adoption strategy
We look at where AI could help and decide what to tackle first. That can be the whole job: a clear plan, with no software to build.
Juan Iturbe / AI consulting & implementation
Panama · Working remotely
Services for businesses.
AI consulting and implementation.
I help businesses use AI to solve problems. We figure out what’s worth doing, choose the tools, and get it working.
Point cloudConnected network
SEED
An animation made with code. Try the controls to change how the points connect.
Technical art × artificial intelligence See my work
01 / What I do
Applied AI Lead
at KetzaiI lead the team building Ketzai Agent. Its agents use each teacher’s context to generate and review educational materials.
My contribution
Ketzai Agent
I lead the team building Ketzai Agent and our work on its architecture, agent coordination, and core runtime.
The product
Educational planning
Teacher Lab helps teachers prepare educational materials. Ketzai has a presence in Panama and the Dominican Republic, with expansion to Colombia planned.
My current role
AI implementation
I work with the team on the tools we use to test agents and check their results.
Monthly figures
- Teachers served
- 4,500+
- Lesson plans generated
- 47,000+
- User growth · July–September 2026
- 13×
My consulting work is separate from my role at Ketzai.
02 / Services
AI consulting and implementation.
We can make a plan, put existing tools to work, test an idea, or build something custom. I also help your team learn to use AI and check its results.
02 What is worth implementing?
Tools and implementation
We connect tools you already use, improve a workflow, or test an idea. If you need something those tools can’t do, we build it.
03 Where should models run?
Models on your infrastructure
Want to run AI on your own servers? We check what fits your data, hardware, and budget, including licenses and ongoing maintenance, then set it up if it makes sense.
My background
From creative pipelines to AI systems.
I worked as a technical artist on productions for Nickelodeon and HBO, connecting tools, workflows, and people to get the work done. Now I do that with software and AI.
Code on GitHub03 / Working notes
What I’m learning about AI.
6 min read AI / Architecture / TS / LangChain
AI Memory Layer: Designing Persistent Memory for Agents
What should an AI agent remember? A practical look at scoped memory, human sources, retrieval, disagreements and where a graph helps.
Read article
10 min read AI / Architecture / LangGraph / TypeScript
Operations, Not Replacements: CQRS Reducers for Agent State
When your agent's state gets complex, full replacements break. Learn how to use incremental operations with CQRS reducers in LangGraph.
Read article
4 min read AI / Evaluation / Implementation
How to Evaluate an AI Proposal Before Hiring
What to ask before accepting an AI proposal: simpler alternatives, representative tests, data boundaries, recovery, operating costs, and pilot acceptance criteria.
Read article
Your next step
Tell me what your business needs
Is there something your team wants to improve? We can look at the problem, try tools, or start with your questions about AI.
Start a conversation