Sergio Villanueva-Meyer
BIM/VDC leadership and AI automation, on real jobsites.
I've spent 15 years running building information modeling at enterprise scale, and now build the AI layer on top of it — agents, MCP-Revit integration, and decision systems working on live construction and real-estate operations. Based in Austin, Texas.
15+
years in construction
$1.25B
largest program led
0→50
VDC team built from zero
$2.5M
documented BIM ROI
Work
Selected projects
Grupo Inmobiliario Habita
Chile · Real-estate construction
$1M+/year in avoidable losses identified
BIM 2.0 diagnostic, a purpose-built CDE platform, and an implementation roadmap connecting design, coordination, and field execution to executive decisions.
viahabita.infratek.ai →PDBM Consulting
USA · Aviation BIM/VDC
4,103 interactions processed in 75 days
Commercial-intelligence system (Jarvis/BIMSearch) that detects BIM/VDC opportunities in public procurement signals and turns noise into qualified conversations.
Enterprise BIM track record
Peru
$2.5M documented ROI
Built COSAPI's corporate BIM department from zero (Banco de la Nación, Metro Línea 2, Pan American Games 2019), scaled DVC Construction's VDC team 0→50 across Peru and Chile, and led enterprise BIM on the $1.25B Lima Airport expansion.
Live systems
Things I've built
A sample of production systems built end-to-end with AI — forecasting, BIM/VDC, agents, and domain tools. Each one is live.
World Cup 2026
AI forecasting + live-data platform with a calibrated, publicly graded prediction record.
worldcup.infratek.ai →VIA Habita
BIM 2.0 common data environment for a real-estate developer (INFRATEK × Habita, Chile).
viahabita.infratek.ai →AgroVIA
3D post-harvest intelligence for agriculture — data to decisions.
agrovia.infratek.ai →Central Electoral
Elections intelligence dashboard (Peru 2026).
elecciones.infratek.ai →JaaS
An AI chief-of-staff agent for operations.
jaas.infratek.ai →Voice
AI sales agents that call, qualify and book.
voice.infratek.ai →Asesor
AI commercial automation for real estate.
asesor.infratek.ai →Milo
A boutique custom-development studio powered by agents.
milo.infratek.ai →INFRATEK Market
A catalog of private AI systems for operators.
market.infratek.ai →Writing
Notes & case studies
Short write-ups on building AI systems that ship. Long-form essays in progress.
Forecasting the 2026 World Cup — and publishing the misses
Jul 2026A calibrated prediction system that graded itself in public: Brier score, hits, and honest misses.
BIM 2.0: a common data environment for real-estate execution
2026Connecting design, coordination, and field execution to executive decisions.
MCP-Revit: connecting LLMs directly to Autodesk Revit
2026The AI layer reading and acting on live BIM models via the Model Context Protocol.
About
M.S. Construction Management (UF) · Stanford VDC · B.Arch · Former President, BIM Committee CAPECO · Claude Certified Architect · Speaker, GTC 2026 Lima
Sergio Villanueva-Meyer
Architect. M.S. Construction Management, University of Florida. Stanford VDC certification. Former President of the BIM Committee at CAPECO.
I've spent 15 years making BIM deliver at enterprise scale: built COSAPI's corporate BIM department from zero with $2.5M in documented ROI, scaled DVC Construction's VDC team from 0 to 50+ professionals across Peru and Chile, and led enterprise BIM implementation for the $1.25B Lima Airport expansion.
Today I combine that operational base with applied AI: MCP servers connecting LLMs directly to Autodesk Revit, commercial-intelligence agents processing thousands of interactions for US AEC firms, and decision systems for construction and real-estate operators.
Now based in Austin, Texas — open to full-time roles: AI generalist, forward-deployed engineer, domain-driven design, or VDC/BIM leadership. Available alongside select consulting engagements.
Consulting
How engagements work
INFRATEK LLC remains an active BIM/VDC and AI consultancy. Engagements run in three stages. Pricing shared in private.
Assessment
A diagnostic of your BIM/VDC maturity and data flow — where rework, cost, and schedule risk actually originate, and what's worth automating first.
Pilot
A scoped implementation on one real workflow: a CDE, an AI agent, or an MCP integration — instrumented so the value is measurable, not theoretical.
Managed operation
Ongoing operation and iteration of the systems in production, with the team trained to run them and the metrics reported to leadership.
Get in touch
For roles, projects, or a straight answer: email is fastest.