I do not just advise on technology — I build it.
Multi-tenant platforms, agents with approval governance, RAG in production and custom integrations. A four-rung ladder: diagnose, design, build and transfer — each one decides whether the next is worth taking.
No commitment · Not a sales call · You leave with a prioritized technical map, not a proposal
Don't have yours yet? Take it free here — 6 min →
Do you recognize any of these blockers?
These are the reasons AI projects stall at pilot. None of them is solved by buying another tool.
"We have the data, but it lives across different systems and nobody agrees on which one is correct."
"We bought AI tools and they still are not connected to the processes that actually matter."
"We ran a pilot that worked in the demo, but never figured out how to take it to production."
"If the vendor leaves, we are left with neither the knowledge nor the ability to maintain it."
Choose your entry point
One ladder: diagnose, design, build, transfer. Each rung decides whether the next one is worth taking.
Architecture Review
A deep review of your stack, integrations, data, and critical flows before scaling AI. You leave with a prioritized technical map and a clear decision on what to keep, redesign, or build.
Take the diagnostic firstSystems Blueprint
Target architecture, data model, integration plan, and a phased build plan. On closing it you choose the build tier — nobody signs blind.
Book a callBuild Sprint
The system running in production, with a 60-day warranty and 30 days of hypercare. Your data never leaves your tenant. I do not bill hourly: the price is measured against the team you avoid hiring.
Book a callFractional AI Architect
Ongoing technical leadership to govern what was built: prioritizing architecture, overseeing execution, and bridging business and engineering.
Book a callThe Build Sprint and retainer are scoped to the engagement: the price is measured against the team you avoid hiring — architect, backend, data, DevOps, frontend — not against an hourly rate.
This is what the Architecture Review hands over
A prioritized map of your stack with one decision per finding: keep, redesign, or build — and the effort each one implies.
| Area | Finding | Decision | Effort | |
|---|---|---|---|---|
| Data | Three different definitions of "active customer" across CRM, billing and support | Redesign | 2 wks | P0 |
| Integrations | CRM ↔ ERP synced by manual CSV, twice a week | Build | 3 wks | P0 |
| Governance | No role-based access control over billing data | Redesign | 1 wk | P0 |
| Infrastructure | No staging environment: changes are tested in production | Build | 1 wk | P1 |
| Team | A single person understands the data pipeline | Transfer | ongoing | P2 |
| Use case | Automatic ticket classification: viable, high estimated value | Build | 4 wks | P2 |
Illustrative example. It is not any client: yours is built on your stack, with your systems and your names.
Technical capabilities already proven in production
This is what I can design and build in addition to optimizing your commercial operation: real systems connected to revenue and day-to-day execution.
Multi-agent orchestration with governed approvals
I create workflows where scouting, proposal generation, and human approval coexist before a system takes real commercial actions.
Multi-tenant SaaS architecture
I design platforms where multiple clients operate on the same technical foundation with clear roles, data boundaries, and experience.
AI content generation pipelines
I orchestrate systems that turn expert knowledge into reusable video, voice, and learning assets to scale content production.
Full-stack cloud systems on Next.js, Supabase, and Vercel
I integrate frontend, backend, data, automations, and deployment so sales, marketing, and operations can run on one living system.
Hiring the team vs. not hiring it
A production AI system normally demands an architect, backend, data engineer, DevOps and frontend. This is the real comparison — not an hourly rate.
| Criteria | Hiring the team | ✓ Build Sprint (Ulises) |
|---|---|---|
| Roles required | 4 – 6 people | 1 architect + prebuilt components |
| Time to first deliverable | 3 – 5 months (recruit + ramp) | 12 weeks from kickoff |
| Team cost (6 weeks) | ~$140,000 USD | Scoped to engagement |
| Risk if someone quits | ❌ The project stops | ✅ Continuity and documented handover |
| Warranty on what ships | ❌ Not applicable | ✅ 60 days + 30 of hypercare |
| Commitment | Permanent payroll | Per project |
| Where your data lives | Depends on the vendor | ✅ Always in your tenant |
How I work with you
Most consultants recommend. I recommend and build — ensuring solutions get implemented, not left in a PowerPoint.
Diagnose
Review the stack, data, integrations, and critical flows before proposing anything. You leave with a prioritized technical map.
Design
Target architecture, data model, integration plan, and a phased build plan with estimates.
Build
Build the system and put it in production: data, integrations, infrastructure, and observability. With warranty and hypercare.
Transfer
Documentation, handover, and support until your team can run and evolve the system without me.
Most AI projects do not die because of the model
of AI pilots never reach production
It is not the model: the data, integrations and governance were not ready. That is why the first rung is a technical diagnosis, not a build.
are abandoned over data that was not ready
Gartner projects that projects without AI-ready data will be cancelled through 2026. It is exactly the first dimension the diagnostic measures.
were approved on ROI that was never measured
Signed on a projection nobody revisited after launch. That is why the Blueprint sets the success metric before anything gets built.
What does not fixing it cost you?
The question is not how much more you will sell. It is how much you are paying today to hold together by hand what a system should do on its own.
* This counts only what you are already spending: manual hours and engineering hours keeping brittle integrations alive. It excludes opportunity cost and the errors that work introduces, so the real figure is higher. Compare it against the team you would have to hire to fix it — that is the right frame, not an hourly rate.
Systems in production
Not projections or third-party case studies: this is what is running today, and you can review it with me live.
Platforms in production
XFactor Academy, Omnistar and this consultancy share infrastructure by design: same Supabase, same CRM, same agent engine.
Active automations
Workflows running on n8n on a dedicated VPS: lead scoring, sequences, reminders, CRM sync and operational alerts.
Off-the-shelf connectors
A scheduler on the Google Calendar API, WhatsApp Business Cloud API with per-number routing, and VAPID push. Integrations built, not bought.
Our Differentiating Factor
Strategic AI Integration for Scalable Growth
We do not implement AI for the sake of it. Every solution is aligned with your growth objectives, existing processes, and team reality.
AI-Powered Business Acceleration
Beyond theory. We build and implement real solutions that generate measurable results in weeks, not months.
Data-Driven Strategies for Smarter Decisions
We transform your scattered data into actionable insights. Every strategic decision is backed by solid analysis, not guesswork.
Customized AI Solutions for Maximum Impact
Every company is unique. We design tailored solutions that adapt to your industry, business model, and growth stage — no generic off-the-shelf solutions.
Proven Track Record
Track record of implementing AI solutions that have transformed sales and marketing processes for companies across multiple industries and countries.
Customized Solutions
Tailored strategies and tools that align with your specific objectives, unique challenges, and real business context.
Hands-On Experience
Direct involvement in all phases of the consulting process, from strategy development to execution and optimization. I do not delegate what matters.
"Most AI projects do not fail because of the model. They fail because of data, integrations, and because nobody defined who runs the system the following Monday."
From those who've seen it firsthand
The training experience with Ulises was exceptional. He managed to transform our commercial team's mindset with practical AI tools and processes that were actually implemented. His hybrid approach of consulting and technology building made all the difference.
Follow me on LinkedIn
Frequently asked questions
Depends on the rung. The Architecture Review delivers the prioritized technical map the same day. The Systems Blueprint delivers architecture, data model and a phased build plan in 3 weeks. The Build Sprint puts the system in production in 12 weeks, with a 60-day warranty and 30 days of hypercare.
Yes, with clear terms per rung. Architecture Review: full refund if by the end of the day you do not have an actionable map. Systems Blueprint: 50% upfront, 50% on delivery; if the document lacks architecture, an integration plan and per-phase estimates, the balance is not charged. Build Sprint: 60-day warranty on what shipped plus 30 days of hypercare, with formal change control. Fractional Architect: 3-month minimum, then cancel with 30 days notice.
No. It is a monthly retainer with no long-term contract. You can cancel with 15 days notice. Most clients stay 6-12 months because the results justify continuity — but we don't lock you in. The first month is for validating fit before committing further.
The Architecture Review is one day: we go through stack, data and integrations, and you leave with a prioritized map of what to keep, redesign or build. It is the snapshot. The Systems Blueprint is 3 weeks of design: target architecture, data model, integration plan and a phased build plan with estimates. It is the blueprint. The Review tells you whether there is something to build; the Blueprint defines exactly what and in what order — and it is where the build tier gets chosen.
Most of my clients are B2B SMEs with teams of 5-50 people. They benefit the most: they have the real pain but not the budget for a full-time CRO ($8,000-15,000 USD/month). The fractional model solves exactly that.
Start with the free diagnostic (6 min). It evaluates 6 architecture dimensions — data, integrations, infrastructure, use cases, governance and team capability — and tells you which rung to start on. If you already know there is something to build, go straight to the Architecture Review ($3,500): one day and you leave with the technical map.
Prefer to write first?
Tell me what you are trying to build, or what is breaking. I answer personally, not an autoresponder.