Enterprise AI · Deployed · Not piloted
Most AI initiatives never reach production.
They don't fail on the model. They fail at one of five gates — and by the time anyone notices, the budget is gone. I find which gate you're stuck at, then get you through it.
Alphabet · DARPA · Salesforce · Aramco · Bridgestone · Pearson · Government of Portugal
The framework
Five gates between funding and production
Every AI initiative that dies, dies at one of these. Most teams are working on the wrong one.
Data
Is the data this use case depends on accessible, governed, and good enough today — not after an eighteen-month cleanup project?
Process
Does anyone actually own the process you're automating? Unowned processes don't get fixed; they get worked around.
Governance
Who signs off when the model is wrong? In regulated environments, an unanswered accountability question stops deployment cold.
Talent
Not “do we have engineers.” Do we have someone who can hold strategy, product, and delivery at the same time — or are we paying four people to translate for each other?
Sponsorship
Is there an executive whose own numbers improve when this ships? If not, it dies at the first budget review.
Gates one and two get all the attention. Gates three through five are where the money is actually lost.
Free · No signup to start
Which gate are you stuck at?
Fifteen questions. Six minutes. You get a score across all five gates, where you sit against other organizations that have taken it, and the three things to fix before you spend another dollar.
No sales call attached. If the answer is “you're not ready for AI yet,” it will tell you that.
Your answers are anonymous and aggregated into the annual State of Enterprise AI report. Email only required to receive your results.
Sample output. Your weakest gate is highlighted.
Engagements
Four ways to work together
AI Opportunity Sprint
Where to start, and why the obvious answer is usually wrong. Discovery with the people who own the broken process, a prioritized opportunity map, and a business case your CFO can defend. Ends with a working prototype of the top use case — not a slide about one.
Fractional Head of AI
Your AI executive, two to three days a week. Strategy, roadmap, governance, vendor selection, and direct leadership of the delivery team. For organizations that need the function before they can justify the headcount.
Executive AI Alignment
A working session with your leadership team. What is real, what is vendor theater, what your competitors are actually doing, and where your next three million should go. You leave with a decision, not a deck.
Regulated Industries AI
Banking, payments, capital markets, and public sector. Where a missed control isn't a retrospective item. Murex MX.3, Basel III, FRTB, PCI — delivered inside these constraints, not around them.
Every engagement starts with a 30-minute call. If I'm not the right person for it, I'll say so and point you somewhere better.
Selected work
What this looks like in practice
Client names withheld under NDA. Sector, scope, and outcome as delivered.
Payments platform — LATAM
High volume of inbound support tickets, most of them repetitive, all of them handled by humans. Designed an AI resolution architecture with confidence-gated escalation and PCI-aware data handling — full technical specification and a working proof of concept delivered in under three weeks.
Global CRM vendor — EMEA
Integration assets sitting idle with no commercial model attached. Architected an API-as-a-Product framework across MuleSoft Anypoint — discovery, versioning, access control, and usage-based billing — turning siloed enterprise data into a recurring revenue line.
Capital markets — Murex MX.3
Regulatory reporting cycles running six hours under Basel III and FRTB. Led MRB integration with external risk and accounting systems; automated reporting feeds brought the cycle under one hour, with zero post-go-live incidents affecting trading or risk operations.
Who you're actually hiring
Ricardo Guzmán
AI Strategy Architect · Founder, Tecnoia
Bogotá, Colombia · Remote-first
Ten years in enterprise software, five focused on applied AI. Embedded with Alphabet's applied AI and enterprise platform initiatives since 2020, running portfolios of 15+ concurrent programs across North America, EMEA, Latin America, and Africa.
Work has spanned DARPA, US Government agencies, Salesforce US & EMEA, Aramco, Bridgestone, Pearson, Textron, Michelin, MTN, Econet, and the Government of Portugal — five continents, mostly in environments where compliance is not negotiable.
I compress four functions into one: AI strategist, product manager, delivery lead, technical lead. Most organizations staff those as four people with four calendars and a translation layer between each. Holding all four removes the coordination overhead that kills timelines.
4 roles · 1 operator · 60–70% less coordination overhead
I design the system, prove it with rapid AI-assisted prototyping before anyone commits budget, and direct the engineering teams that harden it.
Remote-first. English and Spanish.
Generative AI Strategic Leader — Vanderbilt · Claude Platform 101 — Anthropic · SAFe 5 Agilist · CSM · CSPO · PSPO I & II
Start with the diagnostic. Or just send me the problem.
Thirty minutes, no deck, no discovery questionnaire. Tell me what you're trying to do and I'll tell you whether it's a real AI problem, a process problem wearing an AI costume, or something you should buy off the shelf.
Contact
Send me the problem
Tell me what you are trying to do. I reply within 24 hours.