AI-First Full-Stack Developer
I build complete products with AI — from idea to paying customer.
9 years running a real business. Today I design and build AI-assisted software already in production: multi-tenant SaaS, audit agents, and conversational agents serving customers 24/7.
Stack I work with
From running a print shop to building with AI
I spent 9 years running Grupo Langamar Impresiones, a family B2B manufacturing business still in operation. That's where I learned to solve real problems with real constraints, not build for a portfolio. I bring that same standard to AI-assisted software today: every product I build has to hold up in production, not just look good in a demo.
- 9 years running Grupo Langamar Impresiones, a family business still in operation
- 3 in-house AI products in production: Camaleón, Matías, and Natalia
- Certified AI Expert — Universidad Isabel I / BIG school, 150 hours
- Founder of GenIA Labs, an AI audit and development agency for SMEs
Experience
From running a real business to building the software that runs it.
Founder — GenIA Labs
January 2026 – Present · Mexico City- Founded a 360° Web Audit, AI Consulting, and Digital Development agency for Mexican SMEs.
- Designed and built Camaleón, a multi-tenant SaaS for health clinics with 50 versioned migrations, 41 automated tests, and compliance with 4 Mexican regulations, in production.
- Built Matías, a web-audit agent with 44 parallel analysis modules, 94 test files, and a mandatory human-review gate.
- Deployed Natalia, a WhatsApp conversational agent running 24/7 with RAG (pgvector) and anti-hallucination guardrails, qualifying B2B leads in production.
Director — Grupo Langamar Impresiones
2017 – 2026 · Mexico City- Led team, client, and supplier management at a family B2B manufacturing business in continuous operation.
- Directed the business's digital marketing strategy.
- Ran the company for 9 years, applying real-world business judgment to every technical and operational decision.
Stack
What I actually build with in production — not an aspirational list.
Frontend
Backend & data
AI
Tooling & deploy
In-depth projects
Three products in production, from architecture to regulatory compliance.
Camaleón
A multi-tenant SaaS platform for health clinics: a universal console with industry "cartridges." The Nutrition cartridge is already in production, on an architecture built to add more industries (e.g. dental) without rebuilding the core.
- versioned migrations
- 50
- components
- 177+
- automated tests
- 41+
- Mexican regulations met
- 4
versioned migrations
components
automated tests
Mexican regulations met
Matías
An internal GenIA Labs agent that audits websites across 15+ dimensions (legal, SEO, GEO, UX, performance, conversion, e-commerce, PWA, accessibility) in a single pass, with evidence and mandatory human review before delivery.
- parallel analysis modules
- 44
- test files
- 94
- documented REST endpoints
- 38
- CI coverage gate
- ≥65%
parallel analysis modules
test files
documented REST endpoints
CI coverage gate
Natalia
A conversational agent that handles WhatsApp for Grupo Langamar Impresiones: qualifies B2B leads, collects quote briefs, and escalates to the human director — live in production, running 24/7.
- running in production
- 24/7
- humanized reply delay
- 2–4 min
- from the director → bot locks out
- 3 msj
running in production
humanized reply delay
from the director → bot locks out
Other projects
More work built with the same standards.
Pide Menú
A restaurant SaaS: a digital menu with configurable options, radius-based delivery zones (geocoding), optional Uber Direct integration, a diner-facing PWA, and LFPDPPP compliance. Subscription model with Pro, Lite, and Café plans.
Grupo Langamar Impresiones
A B2B corporate site for the family print shop: a full technical SEO/GEO strategy (schema markup, atomic answer capsules for AI Overviews, 23 preserved 301 redirects through the migration), 4 buyer personas with specific objections, and a documented conversion funnel with business and technical KPIs.
GenIA Labs
My own agency's site (dogfooding): Next.js with i18n (ES/EN), a real GEO strategy (llms.txt, llms-full.txt, explicit rules for 9 AI bots in robots.txt, Q&A blocks for citation by generative engines), and 3 in-house AI agents documented with explicit limits on what they must not invent.
Looking for someone who actually builds with AI?
Reach out directly. I answer myself.