Facu Villanueva

_

My obsession is simple: help build the best teams. I direct AI systems to make that real — for how I hire, evaluate talent, and operate at scale. Not demos.

Get in touch → LinkedIn
scroll
01 [ABOUT]

I help build the best teams

I started in Human Resources and moved deliberately toward AI engineering — not to become a developer, but because the systems I needed to build great teams didn't exist yet. So I direct AI to build them.

Architecting AI systems that make hiring and team decisions faster and more accurate — orchestration, evaluation loops, audit trails.
Directing AI-built software end-to-end to solve real team-building problems — product thinking, specification and verification, not hands-on code.
Integrating LLMs into real operational hiring workflows, not just demos.
5+ years running real recruitment operations — so I know exactly what makes a team work, not just what looks good on a slide.
02 [PROJECTS]

Proof, not theory

Everything here exists because it makes a team's hiring sharper, faster, or more reliable — not because it looks good in a portfolio.

AI Governance Harness

An operational framework for running autonomous AI agents safely in production: scoped permission boundaries, a kill-switch, versioned decision logs, rollback strategy, and staged promotion gates before any agent runs unattended — because the tools behind a team deserve the same rigor as the team itself.

Agent OrchestrationGovernanceEvals
→ Fork it, replicate it

Hiring, or want to see how any of this actually works? I'll walk you through it.

Contact me →
03 [WHAT-I-BRING]

Know-how, not a service menu

This is what actually goes into building a great team — the operational skill set behind everything above.

Technical Recruiting, IT Market

core specialty

5+ years recruiting IT profiles end-to-end — sourcing, technical screening, interview loops, offers, and onboarding. ~110 recruitment processes in 2025 alone, 200+ interviews conducted, 45 hires delivered across technical, leadership, and business roles for 10+ product and consulting teams. I know how to evaluate a technical profile I don't personally code — through structured screening, the right interviewers, and reading signal from how a candidate reasons, not just keywords on a CV.

AI Governance & Agent Orchestration

core discipline

Designing operational frameworks for autonomous AI agents: permission scoping, kill-switches, versioned decision logs, rollback strategy, staged promotion gates, and cross-provider model fallback so agent work stays safe running unattended.

Spec-Driven AI Delivery

how I build

I don't write code by hand. I direct AI systems to build production software through precise technical specification, staged builds, and rigorous, evidence-based verification — I know exactly what "done" looks like before I delegate it.

Multi-Agent Evals & Cost Discipline

quality control

Evaluation loops that check AI output before it ships, plus cost/risk-aware model selection — fallback ladders and escalation rules instead of trusting a single model blindly.

HR Domain Fluency, Applied

the edge most builders lack

Beyond recruiting: onboarding, employee engagement, and day-to-day HR operations experience shapes how I direct AI systems — I know what actually changes a hiring or people team's daily workflow, not just what looks good in a demo.

Recruitment Automation, in Production

shipped, not theoretical

Automated workflows (Power Automate, Microsoft connectors) already running inside a real hiring operation: email triggers, candidate follow-ups, and recruitment-data traceability across the pipeline.

Data Fluency

in progress

Building working fluency in Power BI and HR analytics to ground AI-directed automation in real operational numbers, not assumptions.

04 [WHY]

Why this isn't just a job to me

I've watched good teams get built by accident and bad ones get built on purpose — by whoever happened to be understaffed that quarter, not by anyone who actually thought about it. That gap is what I care about closing. Everything else on this page is downstream of that.

A team isn't a headcount number, and a hire isn't a checkbox at the end of a funnel. A pipeline is a series of judgment calls — most hiring tools are built to hide that, not support it.
I don't build software to prove I can direct AI. I build it because the tools recruiters get handed were built by people who never sat in the chair, and it shows in what gets missed every day.
Technology serves people, not the other way around. If a system makes a hiring decision faster but worse, it's not done — it's just fast.
I'm not trying to sell you a SaaS on this page. Selling comes later, if at all. What I actually want is to compare notes with anyone building something real — a team, a product, a company — because how you build the team is the ceiling on everything else you build.
05 [SKILLS]

Stack & tools

// AI & AGENT SYSTEMS
Claude / Claude Code opencode Multi-Agent Orchestration Agent Evals & Fallback Design Spec-Driven Development Governance & Audit Design
// AUTOMATION & DEPLOY
Microsoft Power Automate Make Notion API Vercel (Deploy & Config)
// DATA (DEVELOPING)
Power BI HR Data Analytics
// RECRUITING & HR
ATS Platforms LinkedIn Recruiter Sourcing & Pipeline Management Stakeholder Management
06 [CONTACT]

Get in touch