Freelance architect for CIOs and Heads of AI
I take generative AI for medium and large enterprises from the pilot phase to production on Microsoft Azure: secure, governed, measurable and with costs under control.
Book a Free Strategy Session
Carlos I. Pérez-Sechi
GenAI platform, security, governance and FinOps
What I do
01Pilot-to-Production Assessment
02Pre-Committee Review
03Quality, GenAI Platform and FinOps
Results
Results in production, with real users, not in a demo.
Services
Almost every generative AI pilot I see works. What fails is everything else: identity and secrets, networking, platform tickets, the architecture committee, the cybersecurity review, quality that nobody measures and a bill for unaccounted-for tokens. That is what I work on.
I measure the actual distance between your pilot and the landing zone with a 12-dimension gap map, from identity and networking to who owns and supports the code, and propose the shortest route to cover it.
Learn MoreI review architecture, security and compliance before they reach the architecture committee and the cybersecurity team, with Article 50 of the AI Act translated into technical controls, and turn every first approval into a reusable pattern.
Learn MoreBefore promising reliability, you need to know today's score. I set up continuous assessment against a business-approved golden dataset, guardrails and traceability for every model decision.
Learn MoreI design the shared GenAI platform and its operating model: an AI gateway with quotas, cost per use case and architectures that skip the model when code will do. The token bill stops being a surprise.
Learn MoreWe decide which use cases to solve with Microsoft 365 Copilot, which with Copilot Studio and which with Microsoft Foundry, and in what order to tackle them.
Learn MoreBook a 30-minute session and we'll look at what stands between it and production.
Guides
Practical guides and frameworks from my work taking enterprise AI from pilot to production.
October 2026 · 12 pages · Spanish
A practical guide and checklist for taking a generative AI app from pilot to production in a corporate Azure landing zone: the three phases and what to promise in each, a 12-dimension gap map, 20 go-live checks, cost by design, measurable quality, what to bring to security and the steering committee, and AI Act Article 50 by role.
Next step
Book a free 30-minute strategy session or message me on WhatsApp to discuss how custom AI solutions can address your specific business challenges and drive innovation.
